Page 1

22ND OCTOBER 2012, PODBANSKE, SLOVAKIA Avalanche Prevention Center

1972 - 2012 th

anniversary


th

anniversary


ADVANCES IN AVALANCHE FORECASTING, PODBANSKÉ

AVALANCHE PREVENTION CENTER MOUNTAIN RESCUE SERVICE DEMÄNOVSKÁ DOLINA, JASNÁ, SLOVAKIA

STRAŽAN S.R.O. SPECIAL VEHICLES AND EQUIPMENT

NADÁCIA JÁNA KORCA MESTO LIPTOVSKÝ MIKULÁŠ

ZAJO GRAND HOLET PERMON PETER SMREK NANUK SKI SPORTEN GEOTRONICS SLOVAKIA CHIMPANZEE PIEPS TATRATEA PIVNICA ORECHOVÁ

GEOINFORMATIKA.SK / GEOCOMMUNITY S.R.O. FUNDACJA IM. ANNY PASEK


ADVANCES IN AVALANCHE FORECASTING, PODBANSKÉ


ADVANCES IN AVALANCHE FORECASTING, PODBANSKÉ

SESSION 1 - AVALANCHE FORECASTING, DISTRIBUTION OF THE AVALANCHE BULLETINS TO THE PUBLIC ...................... 11 40 years of avalanche prevention and safety in Slovakia Marek Biskupič, Milan Lizuch, Filip Kyzek, Jozef Richnavský, Igor Žiak..................................................................... New Avalanche Bulletin in Switzerland Gian Darms, Kurt Winkler ......................................................................................................................................... Results of the survey about avalanche reports comprehension in Catalan Pyrenees Glòria Martí i Domenéch, Carles García Sellés, J. Pujol, J. Fleta, P. Oller & P. Martínez ........................................... Snow avalanches as natural disaster in Polish mountains as exemplified by the Polish Tatra Mountains Maciej Karzyoski, Anna Fiema, Tomasz Nodzynski ................................................................................................... Avalanche forecasting in Romania Narcisa Milian ........................................................................................................................................................... Geoavalanche - spatial data infrastructure for avalanche awareness warning Francesco Bartoli ......................................................................................................................................................

12 15 16 20 22 26

SESSION 2 - NEW APPROACHES AND TOOLS FOR AVALANCHE FORECASTING ....................................................... 33 Avalanche danger patterns, a new approach to snow and avalanches science Rudi Mair, Patrick Nairz ............................................................................................................................................ Backcountry Risk Analysis for Fracture Depth and Slope Angle David McClung .......................................................................................................................................................... YETI – a software to service the avalanche forecaster Igor Chiambretti, Mauro Valt, Paola Dellavedova .................................................................................................... LABINIX – a tool for regional avalanche danger assessment from meteorological data based on regression Martin Vojtek ............................................................................................................................................................

34 36 38 44

SESSION 3 - MODELLING AND GIT FOR AVALANCHE HAZARD MAPPING .............................................................. 47 Analysis of weather condition on 25th March 2009 over Slovakia and numerical weather prediction outputs for the day of avalanche occurrence in Žiarska valley Pavol Beránek, Martina Sadlopová ........................................................................................................................... Weather, terrain, vegetation and snowpack based avalanche risk model Anna Seres ................................................................................................................................................................ Experiments with remote sensing in the context of avalanche warning and detection Rune Solberg, R. Frauenfelder, S.Ø. Larsen, A.-B. Salberg......................................................................................... Avalanche forecasting system for the forest managers – common aspects Roman Sitko .............................................................................................................................................................. Snow avalanches mapping – evaluation of a new approach Pawel Chrustek, Natalia Kolecka, Yves Bühler........................................................................................................... Snow Avalanche Risk Assessment in Territorial Planning; Case of Settlement Magurka Ivan Mudrop, Jozef Richnavský, Pawel Chrustek, Marek Biskupič.............................................................................

48 49 56 58 61 69

7 TH SUDETO CARPATHIAN AVALANCHE WARNING SERVICES MEETING................................................................ 74 Water saturated avalanches in Krkonoše Mts. Juraj Juráš, Valerian Spusta, Milena Kociánová, Irena Špatenkova, Jiří Pavlásek....................................................

75

Avalanches as a natural factor – influence on the species biodiversity and the landscape Petra Šťastná, M. Kociánová.................................................................................................................................... 78 Winter season 2011-2012 in Italy: Meteomont activity, snow and avalanche report and snow emergency on february 2012 Vincenzo Romeo, M. Fazzini..................................................................................................................................... 80


ADVANCES IN AVALANCHE FORECASTING, PODBANSKÉ

Avalanche rescue mission is a great challenge for every mountain rescuer. Time is life, therefore the rescue has to be fast, efficient, and safe for the rescuers as well. Despite all the efforts of organised rescue, many accidents end tragically. The prevention carries a great importance and it is our duty and responsibility to prevent the accidents. Forty years ago the Avalanche Prevention Center (APC) has been established. Nine people led by Dr. Kņazovický started a very ambitious plan. Many outputs and methods from that period are still used for the prevention and safety. Publications, avalanche forecasts and warning in addition to awareness courses, all these activities of APC are invaluable for the public prevention. Avalanche prevention center is one of the few institutions in Europe which is in charge of avalanche prevention and rescue on national level. In the field of avalanche rescue, staying in touch with up to date knowledge and technology is crucial. APC regularly takes place at the meetings of IKAR, EAWS and SCAWS and then spreads the expertise and recommendations further among the rescuers. Advances in avalanche forecasting will celebrate the forty years of the Avalanche prevention center existence. We have the opportunity to summarise the development and potentially draw the future trends of this very interesting field. I wish you fruitful meeting and pleasant stay in the Slovak mountains!

Lavínová záchranná akcia predstavuje pre záchranárov veľkú logistickú výzvu. Napriek rýchlemu nasadeniu, pohybu v nebezpečnom teréne a veľkým znalostiam a skúsenostiam pri hľadaní a ošetrení zasypaných, sú výsledky ich snaženia vo väčšine prípadov negatívne. Predchádzať takýmto udalostiam je preto našou snahou a povinnosťou. Pred 40 rokmi začalo svoju činnosť Stredisko lavínovej prevencie (SLP) v Jasnej. 9 ľudí pod vedením Dr. Kňazovického začalo pracovať v inštitúcii, ktorá mala v tých časoch veľké a smelé ambície. Niektoré sa naplniť nepodarilo, ale mnoho činností, aktivít a produktov SLP pomáha pri bezpečnosti návštevníkov našich hôr dodnes. Publikácie, lavínové informácie, výstrahy, či kurzy pre verejnosť sú už neoddeliteľnou súčasťou práce Strediska pre verejnosť. Stredisko lavínovej prevencie, ako jedna z mála inštitúcii v Európe, zabezpečuje činnosti aj v oblasti lavínovej záchrany a tiež v oblasti lavínovej prevencie. Sledovanie najnovších trendov a technológií v lavínovej záchrane a prevencii, udržiavanie kontaktov s lavinármi z celého sveta a členstvo v medzinárodných organizáciách (IKAR, EAWS, SKLS) napomáhajú aj pri vzdelávaní ostatných horských záchranárov a udržuje ich odbornosť a znalosti na vysokej úrovni. Konferencia „Pokroky v lavínových predpovediach“ sa koná pri príležitosti 40. výročia založenia Strediska lavínovej prevencie. Je to príležitosť zosumarizovať pokroky, ktoré sa v tejto oblasti uskutočnili, ale aj nastoliť témy a vízie do budúcnosti. Prajem Vám veľa úspechov a príjemný pobyt v našich horách!

Jozef JANIGA Director of Mountain rescue service Riaditeľ Horskej záchrannej služby


ADVANCES IN AVALANCHE FORECASTING, PODBANSKÉ

You are holding the conference proceeding of „Advances in avalanche forecasting― conference. It has been already forty years since the Avalanche preventions center has been established. The beginnings were difficult but exciting. The idea to start with snow and avalanche science in Slovakia was a step into the unknown. As the time has been passing by the Center shifted its focus from solely scientific approach more to the practical tasks and prevention. Since its founding the Center published numerous expert studies, reports and surveys focused on avalanche hazarded assessment and mitigation. The Center also cooperated on the design of avalanche defence and support structures in Slovakia. But its biggest task and challenge is the avalanche prevention. The prevention is immeasurable and often underestimated, but very important. The best accidents are those who never happened. I can claim that despite the increasing number of winter travellers, the trend of avalanche fatalities and injuries does not rise considerably. This is thanks to avalanche prevention which has been done here in the last decades. The effort and work behind the prevention is not so visible. All the forecasters know what it means to collect the snow pack, meteorological and climatologic data and exercising other field work. For almost 18 seasons the Center has been publishing avalanche bulletin according to the international standards. I would like to express words of thanks to all people who were behind the idea of establishing the Avalanche prevention center, currents employees, and the participants of the conference. Big thanks also goes to the authors of the contributions for creating interesting and wonderful programme. Wish you enjoyable stay and happy upcoming winter! V ruke držíte zborník prác z konferencie „Pokroky v lavínovej predpovedi“, ktorá sa koná pri príležitosti 40 výročia založenia Strediska lavínovej prevencie. Je to už 40 rokov čo na Slovensku existuje inštitúcia, ktorá sa systematicky venuje lavínovej prevencii a bezpečnosti na horách. Začiatky boli ťažké, ale o to viac vzrušujúce. Veď výskum snehu a lavín bol na Slovensku krokom do veľkého neznáma. Postupom času sa však Stredisko lavínovej prevencie vypracovalo na uznávanú inštitúciu, ktorá sa okrem vedeckej činnosti čoraz viac venovala aj praktickým úlohám. Publikovaných bolo viacero odborných štúdií a posudkov lavínovej ohrozenosti objektov, lyžiarskych svahov a komunikácií a navrhované boli účinné protilavínové opatrenia. Jednou z najväčších úloh však bola lavínová prevencia. Prevencia ktorá je síce nemerateľná a často nedocenená, no jej význam je nesmierny. Najlepšie nehody sú tie, ktoré sa nikdy nestali. Dovolím si tvrdiť, že aj vďaka lavínovej prevencii je počet ľudských obetí a materiálne škody na objektoch podstatne nižší. Za prevenciou sa skrýva obrovský kus práce v teréne, nespočetné snehomerné a meteorologické merania a pozorovania a tak isto kus práce pri vyhodnocovaní, interpretovaní a publikovaní výsledkov meraní. Máme za sebou už 18 zimných sezón s denným vydávaním informácie o lavínovom nebezpečenstve v súlade s medzinárodnou lavínovou stupnicou. Rád by som poďakoval ľudom, ktorí stáli pri založení Strediska lavínovej prevencie, súčasným zamestnanom a Vám, návštevníkom konferencie. V neposlednom rade patrí veľká vďaka autorom príspevkov za vytvorenie zaujímavého programu. Želám Vám prijemný pobyt a radostnú nadchádzajúcu zimu!

Marek BISKUPIČ Director of Avalanche prevention center Riaditeľ Strediska lavínovej prevencie


ADVANCES IN AVALANCHE FORECASTING, PODBANSKÉ

Sunday, 21nd October 2012 18:00 - 20:00 Registration

Monday, 22nd October 2012 8:00 - 9:00 Registration 9:30 9:40 9:40 10:00 10:20 10:40 10:50 11:10 11:30 12:00 13:30 13:30 13:50 14:10 14:30 14:50 15:10 15:30 15:30 15:50 16:10 16:30 16:50 17:10

Opening

Session 1 - Avalanche forecasting, Distribution of avalanche bulletins to the public chair: Patrick Nairz

40 years of avalanche prevention and safety in Slovakia Marek Biskupič, Milan Lizuch, Filip Kyzek, Jozef Richnavský, Igor Žiak New Avalanche Bulletin in Switzerland Gian Darms, Kurt Winkler Results of the survey about avalanche reports comprehension in Catalan Pyrenees Glòria Martí i Domenéch, Carles García Sellés, J. Pujol, J. Fleta, P. Oller, P. Martínez Coffee break Snow avalanches as natural disaster in Polish mountains as exemplified by the Polish Tatra Mountains Maciej Karzyoski, Anna Fiema, Tomasz Nodzynski Avalanche forecasting in Romania Narcisa Milian, A. David, A. Nagy Geoavalanche - spatial data infrastructure for avalanche awareness warning Francesco Bartoli Lunch

Session 2 - New approaches and tools for avalanche forecasting Chair: Marek Biskupič

Avalanche danger patterns, a new approach to snow and avalanches science Rudi Mair, Patrick Nairz Backcountry Risk Analysis for Fracture Depth and Slope Angle David McClung YETI – a software to service the avalanche forecaster Igor Chiambretti, Mauro Valt, Paola Dellavedova LABINIX – a tool for regional avalanche danger assessment from meteorological data based on regression Martin Vojtek „Moarri“ – open source software project for management, transfer and visualization for snow cover profiles Jakub Radlioski coffee break

Session 3 - Modelling and GIT for avalanche hazard mapping Chair: Martin Vojtek

Analysis of weather condition on 25th March 2009 over Slovakia and numerical weather prediction outputs for the day of avalanche occurrence in Žiarska valley Pavol Beránek, Martina Sadlopová Weather, terrain, vegetation and snowpack based avalanche risk model Anna Seres Experiments with remote sensing in the context of avalanche warning and detection Rune Solberg, R. Frauenfelder, S.Ø. Larsen, A.-B. Salberg Avalanche forecasting system for the forest managers – common aspects Roman Sitko Snow avalanches mapping – evaluation of a new approach Pawel Chrustek, Natalia Kolecka, Yves Bühler Snow Avalanche Risk Assessment in Territorial Planning; Case of Settlement Magurka Ivan Mudrop, Jozef Richnavský, Pawel Chrustek, Marek Biskupič

19:00 CULTURAL EVENT, Gala Dinner


ADVANCES IN AVALANCHE FORECASTING, PODBANSKÉ

Nedeľa, 21. októbra 2012 18:00 - 20:00 Registrácia

Pondelok, 22. októbra 2012 8:00 - 9:00 Registrácia 9:30

Otvorenie

9:40

Sekcia 1 - Lavínová predpoveď, Distribúcia lavínových správ pre verejnosť

9:40 10:00 10:20 10:40 10:50 11:10 11:30

moderátor: Patrick Nairz

40 rokov lavínovej prevencie na Slovensku Marek Biskupič, Milan Lizuch, Filip Kyzek, Jozef Richnavský, Igor Žiak Nový lavínový bulletin vo Švajčiarsku Gian Darms, Kurt Winkler Výsledky prieskumu o zrozumiteľnosti lavínových správ v oblasti Katalánskych Pyrenejí Glòria Martí i Domenéch, Carles García Sellés, J. Pujol, J. Fleta, P. Oller, P. Martínez prestávka Snehové lavíny ako prírodná katastrofa v poľských horách - príklad poľských Tatier Maciej Karzyoski, Anna Fiema, Tomasz Nodzynski Lavínová predpoveď v Rumunsku Narcisa Milian, A. David, A. Nagy Geoavalanche - štruktúra priestorových údajov pre verejné lavínové výstrahy Francesco Bartoli

12:00

Obed

13:30

Sekcia 2 - Nové prístupy a nástroje v lavínovej predpovedi

13:30 13:50 14:10 14:30 14:50 15:10 15:30 15:30 15:50 16:10 16:30 16:50 17:10

moderátor: Marek Biskupič

Vzorové situácie lavínového nebezpečenstva - nový prístup vo výskume snehu a lavín Rudi Mair, Patrick Nairz Výška odtrhu a sklon svahu - analýza nebezpečenstva vo voľnom teréne David McClung YETI – program slúžiaci pri lavínovej predpovedi Igor Chiambretti, Mauro Valt, Paola Dellavedova LABINIX – nástroj pre posúdenie regionálneho lavínového nebezpečenstva z meteorologických dát metódou regresie Martin Vojtek „Moarri“ – projekt OpenSource softvéru pre správu, prenos a vizualizáciu snehových profilov Jakub Radlioski prestávka

Session 3 - Modelovanie a GIT pre mapovanie lavínového nebezpečenstva moderátor: Martin Vojtek

Analýza poveternostných podmienok nad Slovenskom dňa 25. marca 2009 a výstupy z numerickej predpovede počasia pre deň výskytu lavín v Žiarskej doline Pavol Beránek, Martina Sadlopová Modelovanie lavínového nebezpečenstva na základe počasia, terénu, vegetácie a snehovej pokrývky Anna Seres Experimentálne použitie diaľkového prieskumu zeme na detekciu lavín a lavínového nebezpečenstva Rune Solberg, R. Frauenfelder, S.Ø. Larsen, A.-B. Salberg Lavínový predpovedný systém pre správu lesov Roman Sitko Mapovanie snehových lavín—zhodnotenie nového prístupu Pawel Chrustek, Natalia Kolecka, Yves Bühler Posúdenie lavínového rizika pri územnom plánovaní, modelový prípad z osady Magurka Ivan Mudrop, Jozef Richnavský, Pawel Chrustek, Marek Biskupič

19:00 SLÁVNOSTNÝ CEREMONIÁL, slávnostná večera


ADVANCES IN AVALANCHE FORECASTING, PODBANSKÉ

Tuesday, 23rd October 2012 9:00

Field trip

12:30

Lunch

14:00

Session 1

14:00 14:15 14:30 14:45 15:00 15:15

Snow and avalanches in Slovakia, winter 2011-2012 Filip Kyzek Presentation of the 2011/12 winter season in the Polish Tatra Mountain Maciej Karzyoski Summarisation winter season 2011-2012 in Poland Jakub Radlioski Snow and avalanche in Romania during 2011-2012 winter Narcisa Milian Snow and avalanche in czech mountains, winter 2011-2012 Viktor Kořízek Winter season 2011-2012 in Italy: Meteomont activity, snow and avalanche report and snow emergency on february 2012 Vincenzo Romeo, M. Fazzini

15:30

Discussion

15:40

Coffee break

16:00

Session 2

16:00 16:20 16:40 17:00

Avalanches as a natural factor – influence on the species biodiversity and the landscape Petra Šťastná, M. Kociánová Avalanche warning service in Ukraine Vasilij Manivčuk EAWS - latest news Patrick Nairz Discussion

Wednesday, 24th October 2012 9:30 9:30 9:50 10:10

Session 3 Glide avalanches in Slovakia Marek Biskupič Atypical avalanche in urbanized area – Zawoja Czatoża case Jakub Radlioski Water saturated avalanches in Krkonoše Mts. Juraj Juráš, Valerian Spusta, Milena Kociánová, Irena Špatenkova, Jiří Pavlásek

10:25

Discussion

10:35

Coffee break

11:00

Session 4

11:00

Summary - results of the meeting

11:20

Next Meeting - where?


ADVANCES IN AVALANCHE FORECASTING, PODBANSKÉ


MAREK BISKUPIČ, MILAN LIZUCH FILIP KYZEK, JOZEF RICHNAVSKÝ, IGOR ŢIAK

ADVANCES IN AVALANCHE FORECASTING, PODBANSKÉ

AVALANCHE PREVENTION CENTRE, JASNÁ, SLOVAKIA

40 years of avalanche prevention and safety in Slovakia Marek Biskupič*, Milan Lizuch, Filip Kyzek, Jozef Richnavský, Igor Ţiak Avalanche Prevention Centre, Jasná, Slovakia INTRODUCTION In general it is difficult to imagine how much work there is behind creating an avalanche bulletin – a message trying to tell winter travelers what still might be in the green zone and what is deep into red zone potentially causing an accident. As avalanche forecasters we know how much effort and data is behind efficient avalanche bulletin. Inspire of the fact that it is sometimes hard to get up very early in the morning, dig the snow profiles in bad weather and analyze huge heap of data, we like the job. Nowadays it is quite common to check the weather forecast, avalanche warnings and conditions prior to starting a hiking or skiing tour in our mountain. Most of us use internet or the ―smart ones―stick with the smart phones. The calls on the avalanche bulletin answering machine have rapidly decreased. Now we have apps for checking the avalanche forecast directly in the field. The way of distribution of avalanche bulletins has changes rapidly for last decades. Also the way we work has changed enormously with computers, mobile phones and electronic devices. Let‘s have a look at how the Avalanche Prevention Center has gone through the history to its current stage. HISTORY From 1924 to 1972 several large avalanche accidents were recorded in Slovakia. The five major avalanches caught 105 people and took lives of 50. After these tragic events Mountain Rescue Service in cooperation with Hydrometeorological Institute decided to release first avalanche warnings in the early sixties. It was not the avalanche bulletin as we know it today. Either there was a warning of avalanches or not. Simple two grade scale was applied. The warnings were distributed by the radio on Thursdays, so the people were informed before the weekend. In the mid sixties first permanent

FIGURE 1. The avalanche prevention center started in brand new building in 1972

avalanche observational site was build up on Chopok in Jasná. The station belonged to the Hydro-meteorological institute and one employee Dušan Hollý was responsible for the systematic avalanche observations. Shortly after, the first book the ―Avalanches―dealing with avalanche hazard was written by L. Kņazovický in 1967. Later on Mountain Rescue Service hired one person - Karol Španik. His tasks involved snowpack observation and avalanche monitoring. In 1956 on Kubinska Hola avalanche swept 56 people and killed 6 of them. Due to this and previous tragedies Mountain Rescue Service decided to establish specialized institute dealing with snow avalanches. Avalanche Prevention Center was officially launched on the 6th of December 1972. Eight people (D. Hollý, L.Milan, Ľ. Huťka, J. Korman, T. Kočtúch, V. Dzúrik, S. Sokol, a Š. Hošala) worked there permanently. The first director

__________________________ * Corresponding author address:

Marek Biskupič Avalanche Prevention Center dom HS č.84 032 51, Demänovská Dolina, Jasná, SLOVAKIA tel: 00421 903 624 664 fax 00421 44 5591 637 e-mail: biskupic@hzs.sk

FIGURE 2. The tragic avalanche hit the settlement of Rybô in 1924, causing 18 fatalities


MAREK BISKUPIČ, MILAN LIZUCH FILIP KYZEK, JOZEF RICHNAVSKÝ, IGOR ŢIAK

ADVANCES IN AVALANCHE FORECASTING, PODBANSKÉ

AVALANCHE PREVENTION CENTRE, JASNÁ, SLOVAKIA

FIGURE 4. First avalanche warning published on the internet in 1998

FIGURE 3. Solar radiation measurements

was L. Kņazovický. The very ambitious aim wasto have scientific institute for the whole Carpathian mountain range similar to the SLF in Davos - Switzerland. Even the building was designed to look the same. The cryolab was built as well but actually was never launched. In 1974 the 6th Avalanche Conference of International Commission of Alpine Rescue (IKAR) was organized in the High Tatras. It was the first time IKAR meeting was held outside of the Alps and former Czechoslovakia became a member. After this event the center started its fruitful publication activity. The book ―Danger of avalanches―by L. Milan was published in 1977. Intensive avalanche mapping and monitoring between 1976 -79 resulted in the first comprehensive map of avalanche paths called the avalanche cadastre. In early eighties first yearbook Snow and avalanches was published. Currently the yearbook is published annually with some brakes between 1995 – 2005. AVALANCHE FORECASTING AND CURRENT ACTIVITIES Since the very first beginning one of the main tasks of the Center was avalanche forecasting and publishing the avalanche bulletins. Originally four grade avalanche danger level scale was used. Level 1 and 2 was just generally announced but levels 3 and 4 were always released with special warnings. The warnings were distributed first by radio and print

media later by TV also. With slight variations the bulletin was published like this until 1993 when five levels avalanche danger scale was adopted. First internet avalanche bulletin was published in 1998 in the form of plain text. Consequently own webpage www.laviny.sk has been created with internationally recognized icons and structure of the bulletin. Current activities are generally broad. The main goal of the Center is to provide efficient avalanche forecast service for the public. For the purpose of avalanche forecasting the Center operates and maintains the network of automatic weather stations in the Slovak mountains. Another very important task is to train and teach the mountain rescuers to be efficient and safe during the avalanche rescue. Recently very fruitful collaboration with the avalanche dog handlers‘ squad has been established. The dog handlers are trained how to perform efficient dog search using GPS. The Center is a member of IKAR Avalanche Commission and communicates the latest recommendations further to the professional and voluntary mountain rescuers. Wide public is not being forgotten either. Annual public awareness courses are organized. The proper trip planning is crucial for safe winter traveling in the mountains. Therefore in the winter of 2011/2012 the Center together with hiking.sk launched public web map service (mapy.hiking.sk) with avalanche paths. The aim of this project is to show people places which are regularly affected by avalanches. It is quite often that during the winter time the avalanche triggering with explosives is necessary. This is done under the supervision of the Avalanche prevention center. In cooperation with the Department of pyrotechnics,


MAREK BISKUPIČ, MILAN LIZUCH FILIP KYZEK, JOZEF RICHNAVSKÝ, IGOR ŢIAK

ADVANCES IN AVALANCHE FORECASTING, PODBANSKÉ

AVALANCHE PREVENTION CENTRE, JASNÁ, SLOVAKIA

special education was established: mountain rescuer – pyrotechnican. These experts are authorized to trigger avalanches by explosives. Besides that the Center carries out avalanche mapping, monitoring and investigation of avalanche accidents. All the major and large avalanches are mapped with accurate GPS (accuracy < 1m). The Center also closely cooperates with Anna Pasek foundation on numerical simulations of large avalanches threatening settlements or frequently visited places. Within the Carpathian Mountain region the Center is leader in the area of avalanche prevention safety and hazard mapping. The milestones of the Avalanche prevention Center 1956: Vajskovská dolina avalanche accident (19 persons caught, 16 died) 1968: Kubínska hoľa avalanche accident (53 persons caught, 6 died) The avalanches first book on avalanches by L. Kņazovický 1964/1965: first professional mountain rescuer with task to observe avalanches (K. Špánik) 1972: Avalanche prevention center founded

1974: IKAR Avalanche conference, former Czechoslovakia became member of IKAR 1976 - 1979: Intensive avalanche mapping resulted in Avalanche cadastre (L. Milan) 1980: Atlas of avalanche paths in Slovak socialistic republic (L. Kņazovický) 1981: first yearbook Snow and Avalanches 1989: Practical handbook for avalanche estimation (J. Peťo)

run–out

1993: Five levels of avalanche danger scale adopted 1998. First avalanche bulletin on internet 2000: First automatic weather station 2003: Avalanche prevention center became the part of professional Mountain rescue service 2005: the founding meeting of Sudeto-Carpathian avalanche warning services in Jasná – Slovakia 2007: 14th European avalanche warning services meeting in Vysoké Tatry – Slovakia 2007/2008: Network of automatic weather station build up 2010: International Commission of Alpine Rescue congress in Vysoké Tatry 2011: Avalanche transceiver test in Jasná 2012: Avalanche airbags test in Jasná

FIGURE 5. The avalanche cadastre and its online representation as webmap service, - tool for trip planning


GIAN DARMS, KURT WINKLER WSL INSTITUTE FOR SNOW AND AVALANCHE RESEARCH SLF

ADVANCES IN AVALANCHE FORECASTING, PODBANSKÉ

New Avalanche Bulletin in Switzerland Gian Darms*, Kurt Winkler WSL Institute for Snow and Avalanche Research SLF, Davos Dorf, Switzerland

ABSTRACT Avalanche forecasts are regarded as an important contribution to avalanche accident prevention. Since the first publication of The National avalanche bulletin in Switzerland in 1945, this product has been revised several times. Now it is time to adapt to new possibilities and behaviours in communication. Therefore, the WSL Institute for Snow and Avalanche Research SLF of Switzerland plans to launch the new designed avalanche bulletin for the season 2012/13. The new products will be designed for internet and smart phones. Products will be published twice a day in the four languages german, french, italian and english. To achieve this goal, the forecast will partly be created by using a phrase catalogue. The new avalanche bulletin will consist of an interactive map and a text describing snowpack and weather conditions. This interactive map and the fact, that all products are available in four languages make information much easier available for people from foreign countries. A print version of all products will be provided. Apart from the avalanche bulletin, the completely revised App ‗White Risk‘ will also provide extensive knowledge on avalanche prevention. This makes the ‗new bulletin‘ a powerful information and education platform.

__________________________ * Corresponding author address:

Gian Darms WSL Institute for Snow and Avalanche Research SLF Flüelastrasse 11 CH-7260, Davos Dorf, SWITZERLAND tel: +41-81-417 0111 fax +41-81-417 0110 e-mail: gian.darms@slf.ch


G. MARTI, C. GARCÍA, J.PUJOL, J. FLETA, P. OLLER, P. MARTINEZ

ADVANCES IN AVALANCHE FORECASTING, PODBANSKÉ

INSTITUT GEOLÒGIC DE CATALUNYA

Results of the survey about avalanche reports comprehension in Catalan Pyrenees G. Marti, C. García, J. Pujol, J. Fleta, P. Oller, P. Martínez Institut Geològic de Catalunya ABSTRACT During 2009-2010 and 2010-2011 winter season the IGC issued an on-line survey to assess user's understanding of IGC's avalanche bulletins (AB). The survey consisted of thirteen questions on icon based (ABI) and text based (ABT) avalanche bulletins. We also included at the bottom a brief space for comments and demands. A total amount of 180 answered surveys were recorded during this two seasons. Preliminary results show that in general the user appreciates ABT because of the easiness of viewing and retaining information through icons. However there is a remarkable number of respondents who also considered very important the detailed information of the ABT. Furthermore results show that there is a very large demand of graphic information such us maps representing recent snowfalls, snow depth, avalanche recent accidents and other issues. INTRODUCTION From the beginning of avalanche forecasting at the end of 80's until year 2008 the avalanche bulletins in Catalan Pyrenees have been textbased in terms of regional forecast (Gavaldà and García, 1996).From 2008-2009 season the avalanche team of the Geological Institute of Catalonia, in addition to the pre-existing text based avalanche bulletin (ABT), began to issue an icon-based avalanche Bulletin (ABI). The main objective was to avoid language barriers and at the same time to issue a friendly user oriented bulletins so that the information could be better retained (Martí et al, 2009). ABI has been thought to contain hierarchical information consisting of three tiers and 8 types of icons. ABT has been considered the last (but not the least) tier where avalanche skilled users and professionals could get further and detailed information. In order to assess the general public degree of acceptance of AB, an on line surve y on the c om prehension an __________________________ * Corresponding author address:

suitability of AB was uploaded at the IGC website. AVALANCHE BULLETINS CONTENTS As introduced before, in order to assess avalanche hazard, there are two types of layouts of avalanche bulletins: ABT and ABI for seven different regions in the Catalan Pyrenees. The classical ABT consists of a plain text containing: Danger level rating Avalanche danger assessment Snowpack distribution Meteorological Forecast Tendency of danger level for the next 48 and 72 hours ABI consists of a hierarchical structure describing the following items (figure 1): First tier: danger level rating for the whole Catalan Pyrenees Avalanche danger assessment Second tier: map with danger level icon and brief description Western/Eastern Pyrenees Third tier: six/seven icons showing the main avalanche hazard features: type and size of avalanche, height and aspect distribution of hazard, loading or natural release, time prone wet avalanches, likelihood, and hazard tendency. At the bottom of this tier there is the meteorological forecast provided by the Meteorological Service of Catalonia represented by 4 icons (weather, temperature, wind, snowfall height) Finally as a fourth tier there's the ABT described above THE QUESTIONNAIRE The enquiry is an on-line web enquiry upload at the IGC website (http://www.igc.cat/web/es/ allaus_butlleti_enquesta.html). It consisted of 13 questions distributed in five sections. The last section was intended to collect comments and suggestions. We could summarize it as follows:

Glòria Martí i Domenéch

In the first section we user contact information: Name, e-mail and age.

Institut Geològic de Catalunya Balmes, 209-211, E-08006 Barcelona, SPAIN tel: +34 93 5538430 fax: +34 93 5538440 e-mail: gmarti@igc.cat

The second section consists of seven questions about which information are consulted and needed from the user point of view. They are:


G. MARTI, C. GARCÍA, J.PUJOL, J. FLETA, P. OLLER, P. MARTINEZ

ADVANCES IN AVALANCHE FORECASTING, PODBANSKÉ

INSTITUT GEOLÒGIC DE CATALUNYA

8. Do you consider the comprehension of the icons: none/ low/ enough/ high Data interpretation and memorization habits are a single section where the user gets the next questions: 9. Which avalanche danger level do you consider more when planning an excursion? In this question we only give a random list of avalanche danger level description (without the number) and specifically in this order: very high, low, considerable, and high. 10. Did you consult the legend of icons and the danger level to answer the question above and the icons comprehension? 11. How often do you consult AB's? 12. Which media do you use? 13. Which kind of winter sport do you practice? Finally we included a text box for comments and suggestions.

FIGURE 1. Legend of icons used in ABI.

1. Which information do you search in AB? The answers consisted of a check box with several possibilities: Avalanche danger level, kind of avalanche, height and aspect that could be affected by avalanches, kind of release, time prone to avalanches, all the mentioned options above. The next following questions are radio (yes/no) buttons Avalanche danger assessment 2. Is there some missing data in AB that could help for its comprehension? In case of "yes" a text area is displayed so that the user could specify them. 3. Is the danger level intuitive enough to assess the snowpack stability? 4. Do you consult ABT, ABI or both? 5. Do you think that the icons of ABI are easy to keep in mind when going to an excursion? 6. Do you think that the information of ABT is easy to keep in mind when going to an excursion? 7. Are the seven nivometeorological regions intuitive enough? The whole third section was question 8 that consisted of a list of 8 icons (only the picture) that has to be assessed according to its comprehension. The icons are kind and size of avalanches, Elevation and aspect, additional load, natural release, time prone to avalanches and snowpack distribution.

PRELIMINARY RESULTS AND DISCUSSION All the sections and questions pointed above were intended for knowing the weak and strong points of AB and so to change or design new strategies to offer a better understanding Answers to the first section and to the questions 11, 12 and 13 were used to define the respondent. The majority of users were men (92%), the range of age was mainly between 31-40 years old (39%) followed by 2130 and 41-50 (both 23%), Higher than 50 years old (12%) and less than 21 years old (3%).Concerning winter outdoor activity the 39% were ski mountaineering practitioners, 17% snowshoers, and 15% off-piste skiers. The rest was surfers, climbers and ice-climbers (figure 2). Usually avalanche information is consulted through the web (68%) followed by specialized forums (15%). It is notorious that telephone speakers only represent less than 1 %. In order to plan an excursion the 45% of the users consult the AB daily three days before the activity, the 31% daily during the week before and only the 6% the day before the excursion. As a suggestion some of the respondents asked for icons or graphics showing the previous avalanche danger level in the current AB or its evolution during the week (figure 3). Concerning consulted data and needs we used questions 1, 2, 4 and 12. The 83% of the respondents consult both ABT and ABI, the 13% only ABI and the 4% only ABT (figure 3). Despite ABI seems to be the most consulted one some users consider very important the


G. MARTI, C. GARCÍA, J.PUJOL, J. FLETA, P. OLLER, P. MARTINEZ

ADVANCES IN AVALANCHE FORECASTING, PODBANSKÉ

INSTITUT GEOLÒGIC DE CATALUNYA

FIGURE 2. Information on avalanche respondents.

ABT and asked for more links of ABT in the IGC website. When asking for the consulted issues the 41 % consult all the data followed by the danger level (20%).

ABT and their comprehension we took into account answers to the questions 3, 5, 6, 7, 8, 9, 10. Specifically the order of the questions 8, 9 and 10 was not random but intended to detect whether the respondent memorized and understood correctly. Furthermore question 8 consisted of four radio buttons in order to avoid an odd number of possible answers and so to force the respondent to give a "positive" o "negative" assessment of icons by an even number of possibilities. Results show that icons are well accepted. In general the 83% consider all the icons understandable, specifically the most comprehensible was the icon on aspectelevation (89%) and the less the one about kind and size of avalanches (74%). Concerning avalanche danger scale when planning an excursion the user is more sensitive to Considerable (58%) followed by high (26%), very high (10%), moderate (2%) and low (1%). There was a 3% that considered none of them. As said before after those questions we asked if they had to consult legends: the 27% had to consult the icon legend and the 12% the avalanche danger scale (figure 4).

FIGURE 3. Avalanche information consulted and frequency of consults.

The 71% considered that there's enough information in AB while the 29% consider that should be more for example: avalanche danger levels using maps, steepness, recent avalanche activity and accidents, recommendations for non skilled users, danger level in function of aspect, info about less endangered areas. To analyze the user assessment of icons and

FIGURE 4. Avalanche icons and danger level understability.


G. MARTI, C. GARCÍA, J.PUJOL, J. FLETA, P. OLLER, P. MARTINEZ

ADVANCES IN AVALANCHE FORECASTING, PODBANSKÉ

INSTITUT GEOLÒGIC DE CATALUNYA

Regarding memorization during outdoors winter activities the 75% state that icons are kept in mind, the 20% no and the 5% no answer/do not know (NA/NK). The result for the same issue for texts in ABT were that the 69% memorized the texts the 27% no and the 4% NA/NK. As a general comment the result shows that not surprisingly icons can be memorized better than text. Although the results show a very good comprehension an even memorization, both texts and icons, an in situ questionnaire should be performed to achieve a more satisfactory approach to the user's reality. As an example of field survey there's been an attempt in the Tavascan region (Western Catalan Pyrenees) with fair results (Montoliu 2011 oral communication). Despite icons are intended to avoid language barriers, some French users asked for translation of the whole texts of IGC's bulletins. Thus and in addition to user's comments pointed above, the ABT seems to be fairly appreciated. Since our bulletins were more oriented for civil protection purposes than for recreationists, some respondents asked for a more pedagogic text in the suggestion section. The last two seasons we adapted brief texts for ABI but further efforts will be performed to suite ABT to recreationists in the coming seasons, as other avalanche warning services already do. Especially, because of the great amount of open suggestions (63%) the survey should be considered as a communication tool between the general public and technicians, therefore the survey allows to scope the degree of user's satisfaction ACKNOWLEDGEMENTS The authors wants to express their grateful to all the respondents of the survey specially for their support comments and suggestions, to the Meteorological Service of Catalonia and to all the Geological Risk Section of the Geological Institute and in particular to Núria Bagués, Aline Concha and Anna Rodríguez for helping us with this task. Special thanks also to the staff of the Tavascan ski resort, in particular to Julià Montoliu who encouraged and inspired us with their on site survey.

REFERENCES Gavaldà, J. and García, C., 1996. La predicción del peligro de aludes en el Pirineo de Catalunya. In: Chacón, J. and Irigaray, C. (Editors), Proceedings of VI Congreso Nacional y Conferencia Internacional de Geología Ambiental y Ordenación del Territorio, Granada, Spain, 22-25 April 1996, pp. 605-614 Martí, G, Pujol, J., Fleta, J.; García, C., Oller, P., Costa, O. & Martínez, P. A new iconographic avalanche bulletin for the Catalan Pyrenees: a beginning for a future avalanche forecasting database. Proceedings of International Snow Science Workshop. Davos (Switzerland), September 2009 pp. 361-365.


ANNA FIEMA, MACIEJ KARZYŃSKI TOMASZ NODZYŃSKI SECTION OF NIVOLOGY INSTITUTE OF METEOROLOGY AND W ATER MANAGEMENT

ADVANCES IN AVALANCHE FORECASTING, PODBANSKÉ

Snow avalanches as natural disaster in Polish mountains as exemplified by the Polish Tatra Mountains Anna Fiema, Maciej Karzyński*, Tomasz Nodzyński Section of Nivology, Institute of Meteorology and Water Management, Kraków, Poland INTRODUCTION An avalanche is a mass of snow, which detaches and moves down the slope of the mountain. We can also describe this phenomenon as an imbalance in the snow cover, which makes the mass of snow moving with a certain speed on surface under the influence of slope own weight.

increasing security depends on improving the processes of collection, transmission and processing of data, the introduction of modern forecasting techniques, but most of all from the depth of knowledge of the rules governing this dangerous phenomenon. RESEARCH The aim of the research is to enrich the existing knowledge on the processes of snow that have a direct impact on the formation of avalanches, as well as the development of topographic data, which are used in the infrastructure planning in mountains areas, and will increase security in tourist traffic in the winter. Snow profiles Details of the structure and stability of the snow cover measurements provide stratigraphic profiles. Selected locations represent different and sometimes extreme conditions of snow. Thanks to ongoing research is possible to estimate areas with smaller and greater probability of the occurrence of avalanches.

FIGURE 1. Snow avalanche in Tatra Mountains

Snow avalanches are one of the natural hazards. In addition to the direct threat to human life and health, cause changes in the environment, destroy buildings and tourism infrastructure. Range of occurrence is very local in nature, restricted to mountain areas. In Poland, due to constant and rapid development in recent years tourism and skiing, this problem becomes particularly important. Avalanches are a serious danger to the ever-increasing number of people staying the mountains. The e f f e c t i v e n e s s of m i n im i zi n g l o s s e s a n d __________________________ * Corresponding author address:

Maciej Karzyński Instytut Meteorologii i Gospodarki Wodnej Observation and Measurement Service Department ul. Piotra Borowego 14 30-215 Krakow, POLAND tel: +48 12 6398132 e-mail: maciej_karzynski@imgw.pl

FIGURE 2. Example of snow structure chart

Implementation of snow profile includes: stability of snow (ram profile) determine the nature of the surface layer separation of layers in the snow cover measure the hardness of snow determine the form and size of the crystals measuring the moisture content of snow


ANNA FIEMA, MACIEJ KARZYŃSKI TOMASZ NODZYŃSKI SECTION OF NIVOLOGY INSTITUTE OF METEOROLOGY AND W ATER MANAGEMENT

snow density measurement temperature in the layer cohesion measurement Snow measurement data are entering to the computer program GELINIV, which stores and presents them in graphical form. These data are the input material for the development of the degree of avalanche forecasts and archived data are essential for studies and publications. Registration of avalanches The issue of registration of avalanches is

ADVANCES IN AVALANCHE FORECASTING, PODBANSKÉ

FIGURE 3. Example of avalanches directory

critical in determining both the avalanche paths and determining the active avalanche areas. There is also a feature that allows a continuous review of the hazards of avalanche forecasts. Maps of avalanche activity areas Thematic maps "Snow avalanches in the Polish Tatra Mountains" was created using historical data. Subsequently, these data were compared with actual data, which allows to determine both a potential avalanche areas in the Polish Tatra Mountains, and the trend of the time.

FIGURE 4. Areas of avalanche activity in the Polish Tatra Mountains (such as compilation of historical data and data from two winter seasons - 2005/06 and 2006/07)


N. MILIAN, A. DAVID, A. NAGY NATIONAL ADMINISTRATION OF METEOROLOGY, SIBIU MOUNTAIN RESCUE SERVICE, SIBIU ANSMR, ROMANIA

ADVANCES IN AVALANCHE FORECASTING, PODBANSKÉ

Avalanche forecasting in Romania 1

N.Milian1,* A.David2, A.Nagy3 National Administration of Meteorology, Sibiu, Romania 2 Mountain Rescue Service, Sibiu, Romania 3 ANSMR, Romania

AVALANCHE CASES IN ROMANIAN CARPATHIANS Though avalanches are one of the most spectacular and destructive hazards that cause every year economic damages (mainly forest) and human victims, until 1990, in Romania snow avalanche cases were mentioned only in certain specialized works, mainly as a geomorphic process and sometimes as a risk phenomena. One of the first avalanches mentions in the Romanian Carpathians is the one from april 1702 (or 1704) in Ceahlău Mountain, when the avalanche stroke the Sihăstria monastery and killed twenty monks (Bălan, 2001). Since tourism has developed over the interwar period and mountain clubs have been established (Romanian Alpine Club, Touring Club, SKV), several accidents have been reported. But tourism development has also led to the first avalanche accidents involving tourists, skiers or climbers. Most of the accidents have been reported in the Făgăraş and Bucegi Mountains, but they have occurred in all the Carpathians massifs. First studies about avalanches have been made in 1963 and 1964, when many avalanches happened all over the Carpathians, blocking railways and highways: 150 avalanches in the Maramureşului Mountains, about 30 avalanches in Rodnei Mountains and 20 in Bihor Mountains. After a government decision was released, forest districts have inventoried avalanche corridors (Gaspar et al, 1968). After 1990, avalanche cases have been recorded at the Faculty of Geography of the Bucharest University and the Department of Geography of Timişoara West University. The activity was improved after 2000. After the avalanche in Bucegi Mountains from February 2004 when four skiers were killed, the __________________________ * Corresponding author address:

Narcisa Milian National Meteorological Administration Transilvania-Sud Regional Forecasting Center 49 Somesului str. 550003 Sibiu, ROMANIA tel: +040742177983 e-mail: narcisa.milian@gmail.com

National Meteorological Administration started a program for snow and avalanche monitoring in Bucegi and Făgăraş Mountains. The Nivometeorological Programme was set in partnership with Météo France, Centre d‘Études de la Neige-Grenoble, for the study of snow, its future evolution and avalanche triggering conditions. The methodology uses classical meteorological observations, specific snow observations and the profile of snow layer resistance. All the data are analyzed using two systems developed by Centre d‘Études de la Neige-Grenoble: GELINIV and CROCUS-MEPRA PC. Using informations from Sibiu and Braşov Mountain Rescue Service, papers or articles (www.dinumititeanu.blogopedia.biz), literature (Voiculescu, 2002), and from the National Administration of Meteorology database, we made a statistics in decades about the avalanche cases with victims and fatalities (fig.1).

FIGURE 1. Avalanche cases,number of victims and fatalities in the Carpathians

The figure shows that very many cases have been reported after 2001, after the beginning of snow and avalanche monitoring program within the National Administration of Meteorology in february 2004 and as a response to the increased request and interest on the avalanche activity. Most of the avalanches with victims happened in the Făgăraş Mountains, followed by Bucegi, Ceahlău, Rodnei - fig.2. The number of reported avalanche cases was taken into consideration, as well as the number of persons caught by avalanche and the deceased. Though in the Bucegi Mountains there are more


N. MILIAN, A. DAVID, A. NAGY NATIONAL ADMINISTRATION OF METEOROLOGY, SIBIU MOUNTAIN RESCUE SERVICE, SIBIU ANSMR, ROMANIA

tourists and skiers, because the access is easier, the number of avalanche cases and victims is higher in Făgăraş Mountains. An important role to this classification plays the most dramatic avalanche in Romania, that happened at Bâlea-Lac in 17th april 1977, when twenty-three people from Sibiu died in an avalanche, among them sixteen children. Also, during the construction of the Transfăgăraşan Road in Bâlea Valley, many workers and soldiers died being caught in avalanches. After the tragic event from april 1977, on January 1, 1979, the meteorological station in Bâlea-Lac started the activity. The number of people caught by avalanches or even deceased in other Romanian mountains is much smaller (up to 4) – fig.2.

FIGURE 2. Avalanche cases, victims and deceased by massifs; known data until June 2012

Using the same data mentioned above, a statistic of the valleys where avalanches with victims where reported can be made. Thus, in Făgăraş Mountains, most people caught in avalanches accidents were in Bâlea Valley. In Bucegi massif, most people trapped by avalanche were in Morarului Valley The Nivometeorological Program Within the nivometeorological program, predictions and diagnosis of snow layer properties are made, together with snow stability and avalanche risk estimations. Daily bulletins are provided, including meteorological forecast and avalanche risk estimations. The daily snow and avalanche bulletin is published since the 2005-2006 winter and it is delivered to the local authorities – Mountain Rescue Teams, district councils, mass-media, touristic resorts and.on several internet sites, blogs and forums. The observational network includes meteorological stations from mountainous regions high frequented by hikers and skiers: Bucegi, Baiului and Făgăraş Mountains. The stations are situated between 1090 and 2500 m

ADVANCES IN AVALANCHE FORECASTING, PODBANSKÉ

and cover most of the mountain area: Vârful Omu - 2504 m, Sinaia – 1510m (Bucegi), Predeal – 1090 (Baiului) and Bâlea-Lac – 2055m (Făgăraş Mountains.). Between 2006 and 2009 observations have been made also in Postăvaru – 1784m (Postăvaru Massif), until the meteorological station was closed. The observations within the program are made daily and weekly. Daily observations are made twice a day, at 06 and 12 UTC, and includes meteorological data, specific snow characteristics (snow temperature at 10 cm below surface, snow type, snow crystal type at the surface and their medium diameter, depth penetration of the nivological sonde into the snow layer), characteristics of the avalanches occurred in the visible area: number, short description, type, triggering altitude, exposure and a local estimation of the avalanche risk degree. Every weekly measurements concern the snow structure and determine: the resistance of the snow layers, every internal snow layer, with their grain types, density, hardness and humidity, together with every 10 cm of snow temperature. Avalanche risk estimation for a specific mountainous region requires a good knowledge of the area (relief, climate), good weather forecast and a most accurate snow metamorphosis forecast. The avalanche risk is estimated using the European Avalanche Danger Scale. This is a 5 -level risk scale undertaken by the European countries since 1993, and since 1996 by Canada and United States (www.slf.ch/ lawineninfo/). Avalanche risk estimation in the National Administration of Meteorology A comparison between the greatest estimated avalanche risk over the past winters shows that the most used was level 3 - considerable. The greatest avalanche risk is considered for both mountainous regions – Bucegi and Făgăraş. The very high (5) and low (1) avalanche risk were used rarely. The data are taken from the National Administration of Meteorology annual Snow Reports from 2004 to 2011 (fig.3). A higher avalanche risk was estimated in case of heavy or abundant precipitations, as well as the metamorphosis of snow crystals inside the snowpack have formed instable structures that can be favorable to avalanche release. The observed avalanches happened as well to a high (4) and very high risk (5), as to a considerable (3) or moderate (2). The avalanches have been triggered spontaneously or by the skiers, hikers or snowboarders, and


N. MILIAN, A. DAVID, A. NAGY NATIONAL ADMINISTRATION OF METEOROLOGY, SIBIU MOUNTAIN RESCUE SERVICE, SIBIU ANSMR, ROMANIA

some of them were fatal. As to the period of time, avalanches happened all over the winter season, when the snowpack was more consistent, but most of them occurred between february and april. Using the avalanche data gathered since the beginning of nivo-meteorological program, for the monitorized area (Fãgãraş and Bucegi Mountains) most avalanches occurred on march (30%), than in april (21,5%), february and may (13,4%) (fig.4).

ADVANCES IN AVALANCHE FORECASTING, PODBANSKÉ

FIGURE 3. Greatest avalanche risk estimated daily for every winter since 2004 (%), compared with the average values (horizontal lines)

vcitims occurred, related to the forecasted danger level, it can be seen that most of the accidents occurred by a considerable 3 level risk (fig. 5).

FIGURE 5. Avalanche risk cases when people were involved

FIGURE 4. Avalanche accidents from january 2004 to june 2011; data within the nivo-meteorological program in National Administration of Meteorology

Between the two massifs, the avalanche risk estimated for Făgăraş is higher than in Bucegi for every winter. This is due to the fact that in Făgăraş the amount of snow is usually higher than in Bucegi, and the orographical aspects are more favorable to avalanche releases – all the northern part of the Făgăraş mountains is very steep, with glacier valleys over 2000 m. However, in some days, the estimated avalanche risk in Bucegi was higher than in Făgăraş. As the days when avalanche accidents with

Other avalanche studies Besides that program, universities are playing an important role in avalanche studies. Avalanche related hydrologic risk and morphodynamic potential in some southern valleys in Făgăraş and Piatra Craiului Mountains have been studied by a group from the Faculty of Geography, University of Bucharest (Alexandru Nedelea, Laura Comanescu, Anca Munteanu, Gheorghe Catalina, Razvan Oprea). GIS technique and 3D modeling was used for snow avalanche risk mapping in Ceahlau National Park by a group from the Faculty of Geography-Geology, Al. I. Cuza Iaşi University (A. Covăsnianu, I.R. Grigoraş and all).


N. MILIAN, A. DAVID, A. NAGY NATIONAL ADMINISTRATION OF METEOROLOGY, SIBIU MOUNTAIN RESCUE SERVICE, SIBIU ANSMR, ROMANIA

The results were adapted and interpreted considering to the European Avalanche Hazard Scale. This work was made in the context of the elaboration of Risk Map and is directly concerning both in the security of tourism activities but also in the management of Natural Park Ceahlău. Dendrological aspects are considered by professor Mircea Voiculescu and his group from the Faculty of Geography, West University, Timisoara. CONCLUSIONS Avalanches studies in Romania are still at their beginning. It is highly necessary to widen the snow observation network to all the existing meteorological mountain stations, but also to collaborate with other institutions involved in the people and economic safety (Jandarmerie, Urgence Services…). Thus, a daily bulletin could be made for all the mountains with high avalanche risk. Maintaining the rules of the avalanche european danger scale is also important, in order to offer an accurate information for every tourist, from Romania and abroad.

ADVANCES IN AVALANCHE FORECASTING, PODBANSKÉ

All the other avalanche studies are highly important, in order to make risk maps for the endangered mountains. REFERENCES Administraţia Naţională de Meteorologie, Bilanţul nivologic al sezonului de iarnă, 2003-2004 until 20102011, Bucureşti Bălan I. (2001), Patericul Românesc, Mânăstirea Sihastria Covasnianu A., Grigoras I.R., State L.E., Balin D., Hogas S. and Balin I., Mapping snow avalanche risk using gis technique and 3d modeling in Ceahlau Mountain Gaspar R, Munteanu S.A. (1968), Studii privind avalanşele de zăpadă şi indicarea măsurilor de prevenire şi combatere, Analele ICAS Greene E., Wiesinger T., Birkeland K., Coléou C., Jones A., Statham G. (2006), Fatal avalanche accidents and forecasted danger levels: Patterns in the United States, Canada, Switzerland and France, ISSW Milian N, Stăncescu M, Avalanches - extreme winter events. Monitoring and avalanche risk, AERAPA, 2012, Cluj-Napoca Voiculescu M, (2002), Fenomene geografice de risc în Masivul Făgăraş, Timişoara http://www.avalanches.org/basics/degree -of-hazard/, accesed on November, 20, 2011 http://www.dinumititeanu.blogopedia.biz/10, avalanse/2avalanse-catastrofale-statistici, accesed on December, 20, 2011


FRANCESCO BARTOLI

ADVANCES IN AVALANCHE FORECASTING, PODBANSKÉ GEOBEYOND, ROME, ITALY

GEOAVALANCHE—Spatial data infrastructure for avalanche aearness warning Francesco Bartoli* Geobeyond Srl, Rome, Italy

ABSTRACT Avalanches are a serious problem across the Alps even more in the latest years, considering the number of people playing outdoor activities on snow areas. Dissemination of information across those stakeholders involved in all warning and rescue services is crucial at each stage. This paper figures out how people involved in risk mitigation might be allowed to fill this gap and exchange information in a common language both in term of contents and contexts without any kind of misleading. Furthermore, several user-oriented services might be acted on Web channels to reach mountaineers via mobile devices and make their experience safe-effective. GeoAvalanche is an open-source project aimed at sharing information on snow avalanche (such as bulletins, incidents, snow-profile, weather observations…) with a common standard in order to build a Spatial Data Infrastructure for cross-border interoperability and early warning alert systems toward a safe mitigation risk of mountain activities across the Alps. The GeoAvalanche server has the OGC Web Feature Service capabilities to enable common warning services for snow avalanche information exchanges in compliance with CAAML (Canadian Avalanche Association Markup Language) specification (adopted by the avalanche community as an international standard) and map visualization services for incident report and bulletin alert. 1 STATE OF THE ART Environmental risk management is a major scope issued by INSPIRE directive. The theme ―Natural Risk Zones‖ asserted in the Annex III identifies all atmospheric, meteorological, hydrologic, geological and wildfire phenomena that, because of their location, severity, and frequency, have the potential to seriously affect population. Specifically, it defines particular areas with significant snow cover combined with steep __________________________ * Corresponding author address:

Francesco Bartoli GEOBEYOND Via M.Augusta 68 02040 Vacone, Rieti, ITALY tel: + 39 333 2997173 e-mail: francesco.bartoli@geobeyond.it

slopes – amplified by snowdrifts – that are prone to influence the occurrence of avalanches and snow slides. In this context, the investigations and the underlying purposes mainly concern the provision of a suitable Spatial Data Infrastructure (SDI). This would give the opportunity to interoperate with systems aimed at regulating the land use and the resource management in areas under certain restrictions and linked to such risk, and would provide a web mapping of those areas susceptible to snow slides by dividing them into zones according to different risk classes [INSPIRE]. A first example of geographical feature – as tool mostly known by the public – is the bulletin that is issued according to the avalanche danger scale whose risk levels are now commonly accepted and universally recognized by all organizations. This standardization conceives a reference to the mapping of those areas at risk, and allows representing thematic maps with a unique legend that can be understood worldwide. Despite the treat of avalanche bulletins in a map [Nairz, P.] has been recently examined, modern information systems show a weak use of geospatial standards that should be in support of interoperable services among organizations and would consequently lead to a snow avalanche data infrastructure with a solid basis. 2 OBJECTIVES Avalanche Warning Services (AWS) in Europe daily collect meteorological and snow data for their nowcasting and forecasting products. Usually, an avalanche center has to inform the relevant authorities in order to let them immediately take actions and effectively alert recreational users to the possible danger of avalanches in those areas where they are planning to venture (i.e.: ski-touring, mountaineering, snow shoeing, etc.). Therefore the information system of choice must be accurately designed with geospatial standards in mind that lead to interoperability between applications, systems and communication through all possible Web channels [Hervàs, J.]. Consequently it will be easier to be able to achieve decision-makers and third-party alpine service providers as well as end-users with their mobile devices (Smartphones, iOS, Androids). In addition, this would strengthen the network of observation centers because


FRANCESCO BARTOLI

ADVANCES IN AVALANCHE FORECASTING, PODBANSKÉ GEOBEYOND, ROME, ITALY

currently a nationwide fragmentation of the forecast seems to be evident and often the harmonization of the final products is actually weak. This delay is even more accentuated at cross-border level where AWS aren‘t allowed to access real-time data on contiguous area belonging to neighboring countries because a shared information system covering mountain areas is still missing. The relevance of data sharing among avalanche organizations resides hence on the geographical nature of the information, which they publish on daily basis. In order to make the most effective use of data and since their spatial component, it is crucial that AWS have to collect them through modern geographic information systems (GIS) [Magnùsson, M.M.] and geospatial database. Nowadays, these organizations need to structurally treat all information related to their observations and to offer advanced monitoring services for hazard and resource management rather than just avalanche maps. This work intends to propose a solution to the problem through the use of a robust geospatial product that has been further evolved to adapt with the Canadian Avalanche Association Markup Language [CAAML], a commonly used semantic in the snow avalanche domain. The objective of this solution is to improve the current cooperation of avalanche centers and the services that they are providing over the Internet for the general public. 3 GEOAVALANCHE PROJECT 3.1 Scope of the initiative GeoAvalanche is an experimental open source project aimed at designing, developing, and testing functionalities and architectures that establish a wide methodology able to deal with snow avalanche datasets in a common way by ensuring interoperability among cross-border, national and subnational organizations based on geospatial standards in the realm of Open Geospatial Consortium [OGC]. Overcoming the current lack of cooperation and accessible services for the public, this project wants to point out a basic tool for improving collaboration among risk management offices and serving a reliable product for alerting. GeoAvalanche server results in accessing the aforementioned datasets and making them available by using OGC Web Feature Service (WFS). It is also intended to provide maps of all stored geographical features (i.e.: bulletins, etc.) through OGC Web Mapping Service (WMS) in order to publish them on third-party Internet sites and on mobile devices equipped

with geolocalisation services. It would be a core component for an upcoming snow avalanche data infrastructure that benefits from all the main features of the geospatial Web. 3.2 Snow avalanche standards 3.2.1 Avalanche Danger Avalanche bulletins are basic tools providing an overview of the snow cover as well as the state of the snowpack by pointing to the avalanche danger issued in a given territory – according to the weather forecasts and the snow profile evolution – for warning purposes so as to contrast the triggers of avalanches and, therefore, possible incidents. This assessment, whose semantic rules are standardized in Europe by the European Avalanche Warning Service organization (EAWS), is carried out for each region by giving out a shortly text description on the basis of avalanche danger. It also contains the edge of the snow, dangerous places or elevations with critical rose of aspects, and finally a graphical representation in a map showing the color theme related to the danger level and the corresponding text portion [Chiambretti, I.]. Avalanche danger scale is then encoded according to the European wide standard currently used at EAWS and divided into 5 classes (1-Low, 2-Moderate, 3- Considerable, 4 -High, 5 -Very High) related to the relevant safety information provided to the user. 3.2.2 CAAML This semantic is an XML grammar language initially developed in 2003 and currently used by the Canadian Avalanche Association to provide a shared encoding structure as well as the exchange of snow avalanche related information over the Internet. The feature types currently supported by CAAML are the following: – Avalanche incident information – Avalanche activity comments – Avalanche observations – Avalanche bulletins – Avalanche closures – Observations on the field – Snowpack structure comments – Snow profile observations – Weather observations Following the development of Geography Markup Language [GML], the organization of elements in CAAML reflects the objectpropertyvalue model pattern, which encodes the


FRANCESCO BARTOLI

ADVANCES IN AVALANCHE FORECASTING, PODBANSKĂ&#x2030; GEOBEYOND, ROME, ITALY

types in CAAML and then assigns properties to each of them. The latest version 5.0 consists of 9 schema files structurally organized as follows:

FIGURE 2. Schema file structure of CAAML version 5.0

Since the nature of CAAML strictly derives from GML, it was designed with the same flexibility. Actually, this recent version borrows the concept of profile from GML, which allows dealing with a logical limitation of the elements relevant to a specific application while keeping the ability to be validated against the overall CAAML standard. Current experiments investigated a profile suitable among the EAWS agencies for their CAAML-scoped avalanche bulletins.

FIGURE 1. European Danger Scale with Recommendations

This specific schema file maintained at this location:

is

currently

http://caaml.org/Schemas/V5.0/Profiles/ BulletinEAWS/CAAMLv5_BulletinEAWS.xsd

Figure 3 shows how a bulletin element has to be semantically expressed in the European profile. This data type is a kind of complex feature that needs to be further exploited in order to explain how it could be published through an endpoint service.


FRANCESCO BARTOLI

ADVANCES IN AVALANCHE FORECASTING, PODBANSKÉ GEOBEYOND, ROME, ITALY

The present approach doesn‘t mean that GeoAvalanche architecture was designed for a limited set of CAAML profiles but, instead, it means that it was designed to handle any of them. As a result, the GeoAvalanche server would be able to manage the exchange of any profiled elements and, hence, to achieve interoperability at different levels (regional/ national/european).

FIGURE 3. XML schema definition for CAAML bulletin type asserted in the EAWS profile

3.3 GeoAvalanche Server GeoAvalanche server is built upon GeoServer [GeoServer] then equipped with its plug-in for supporting third-party GML application schemas. This latter functionality allows serving complex snow avalanche features encoded by CAAML. The project is developed under GNU General Public License v3. 3.3.1 GeoServer GeoServer is the reference implementation of OGC Web Feature Service (WFS) standard, and also supports OGC Web Map Service (WMS). WFSs are of particular interest for data interoperability because, unlike a portrayal service such as WMS, they allow directly querying the underlying data. GeoServer is a powerful geospatial engine able to aggregate different datastores at a single point and to let them be republished as cascaded Web Services from distributed sources including also remote WFS as shown in the following figure 4. As the support for GML 3.2.1 is already developed, GeoServer can comply with the INSPIRE Directive that requires to issue WFS s erv ic es in ac c or d anc e wi t h th e abovementioned GML version.

FIGURE 4. GeoAvalanche server implementation

3.3.2 Snow avalanche application schema support GML application schemas can indiscriminately represent complex information models such as CAAML for snow avalanches. GeoServer application schema support is applicable to CAAML thanks to spatially described information that is represented in complex features expressed as GML 3.2.1 application profile. It is currently maintained as a standard GeoServer plugin, which makes use of the simple feature access provided by GeoTools [GeoTools] and converts each of them – retrieved as database tables – into complex features by using mapping rules. As a single caaml:Bulletin can be observed at several different locations on the Earth‘s surface, it can have one of the multivalued


FRANCESCO BARTOLI

ADVANCES IN AVALANCHE FORECASTING, PODBANSKÉ GEOBEYOND, ROME, ITALY

caaml:bulletinResultsOf properties, each of them being a caaml:BulletinMeasurements. The resulting mapping can be depicted as shown in figure 5. GeoAvalanche deployments also include a spatial DBMS, such as PostGIS, to supplement GIS functionalities for CAAML complex features and, therefore, they become together a good fit for all those features that you might expectfrom a SDI. The GeoAvalanche component within a CAAML data infrastructure plays a key role because it manages both read and write operations regardless of the database schema used to store such data. It can perform WFS filter queries and also acts according to OGC WFS-T transactional specification because each single service is conformed to the same CAAML application schema [Caradoc-Davies, B.]. On the other hand, its service-oriented architecture allows exploiting lightweight format like GeoJSON for consuming data from mobile, custom-client and any third-party system. A straightforward request for bulletins can be expressed as follows: http://localhost:8080/geoavalanche/avy/ows? service=WFS&version=1.0.0&request=GetFeature&t ypeName=avy:bulletins&outputFormat=json

and further refined in order to filter outappropriate macro-zones through a CQL syntax like cql_filter=(res=’Monte Rosa”). 4 SPATIAL DATA INFRASTRUCTURE Interoperability is the first step towards a snow avalanche data infrastructure where distributed observation centers and central avalanche warning services can mutually exchange them through a shared semantic structure. GeoAvalanche server offers a wide flexibility in building a nationwide network of regional AWS departments. In fact, once the national authorities decide to leverage regional offices to using CAAML, GeoAvalanche will play a key role in setting up a new prospect of data-driven services. This paper is highlighting the approach on the data exchange rather than their visualization. This is regularly possible thanks to the maturity of GeoServer that offers default outof- the-box capabilities to display maps via WMS standard implementation and,

FIGURE 5. Simple feature mapping of the CAAML complex feature

hence, that will lead to easily deliver thematic maps of their nowcasting and forecasting products. Actually, each single office could potentially implement its own CAAML-based server and act either as a remotely accessible node that simply collects observations, or as a regional warning service that provides the end users with final local products. The relevance of this methodology is mainly its ability to compose aggregated maps rather than limited and to easily support the integration of existing CAAML tools in a transparent way. As a result, our proposal for a preliminary European CAAML Spatial Data Infrastructure is issued as shown in the figure 6. This architecture relies on a distributed serviceoriented paradigm essentially based on the underlying cutting-edge technologies for the GeoWeb by strengthening the use of CAAML and making GeoAvalanche the basic building block for the Web 2.0 tools across the snow avalanche community. By using these integrated Technologies, forecasters and officers will be able to respond to several risks (avalanches, hydropower resource, technological) by using the following tools: National integrated platform to manage distributed sources of snow avalanche observations through an interoperable mechanism; Avalanche Bulletin Map widget, at either national or subnational level, which retains maps of nowcasting and forecasting and other relevant information (incident reports, snow cover, snow quality, etc.) with the possibility to access third-party services for data fusion capabilities; On-demand Warning Services by using RESTful endpoints for retrieving reports of snow depth, daily, weekly and monthly snow water equivalent, and more specifically located alerts about bulletins, main closures (highways, local roads, ascents), severe weather conditions, etc.;


FRANCESCO BARTOLI

ADVANCES IN AVALANCHE FORECASTING, PODBANSKÉ GEOBEYOND, ROME, ITALY

FIGURE 6. Proposal for a European snow avalanche Spatial Data Infrastructure

Transnational data catalogue for discovering data and services about snow avalanche information and consequent resources (water, energy, etc.) with a shared legend and glossary; European Avalanche Map Composer for producing cross-border maps and collapsing national hazards into simple view of governance tools.

From an application viewpoint, the use of its WFS services would achieve an interoperable network among either EAWS agencies or liaison regional offices, and would further leverage new improvements to develop innovative services of early warning systems which take into account the safety of backcountry tourists as well as some qualitative information on the snow. The solution would be able to obtain:

5 RESULTS AND CONCLUSIONS The system was easily prototyped and tested to focus on a simple CAAML datastore in order to effectively demonstrate the feasibility of such experimental SDI. However, the results have to be locally validated with reference to a real environment so as to centrally collect – at a national avalanche agency – data from distributed sources and then to publish avalanche bulletin maps of the underlying regions. The development of an SDI among avalanche centers (and other possible stakeholders) based on GeoAvalanche will ensure data sharing, interoperability, and more accurate information for nowcasting and forecasting purposes.

At European level: a snow avalanche geoportal collecting data from each member state and representing a contribution to define natural risk zone to INSPIRE; At National level: a geospatial tool providing avalanche bulletin maps, value added location-based services to mobile users such as alerts, and finally a complement for resource management by taking into account the snow water equivalent; At Regional level: a measure integrating the observations collected from the ground in a format easily shared, commonly agreed, and useful to a single national container for such information.


FRANCESCO BARTOLI

ADVANCES IN AVALANCHE FORECASTING, PODBANSKÉ GEOBEYOND, ROME, ITALY

6 FUTURE WORKS Despite the outcomes confirmed the ability to exchange snow avalanche datasets with a standard common language, more challenges can be raised. Actually, the upcoming implementation of services requiring compliance with the INSPIRE data themes needs to be further investigated and a schema transformation for CAAML datasets needs to be defined. Furthermore, interoperability has to be complemented by catalogue functionalities so as to offer search capabilities on avalanche metadata and, thus, to improve the impact on the user experience. Therefore, future works will be focused on developing Catalogue Service for the Web (CSW), which will enable extended WMS configuration for serving relevant WMS 1.3 bulletin alerts to map visualization services that comply with INSPIRE. REFERENCE CAAML, Canadian Avalanche Association Markup Language, http://caaml.org Caradoc-Davies, B., Angreani R., 2010, GeoServer Application Schema Support: Complex Web Feature Service for Geoscience Interoperability, 4th eResearch Australasia Conference, Gold Coast, Australia, CSIRO Earth Science and Resource Engineering Chiambretti, I., Bartoli, F., 2010, Gestione del rischio valanghe e sue applicazioni XML, GEOmedia n° 3-2010, 14-17, Roma, Italia, Rivista Geomedia GeoServer, http://geoserver.org GeoTools, http://geotools.org GML, Geography Markup Language, http://www.opengeospatial.org/standards/gml Hervàs, J., 2003, Recommendations to deal with Snow Avalanches in Europe, NEDIES project, 71 -75, Ispra (VA), Italy, Joint Research Centre. INSPIRE, Directive, Drafting Team "Data Specifications", 2008, D2.3 Definition of Annex Themes and scope v3.0, 100-101, INSPIRE Infrastructure for Spatial Information in Europe, European Commission Magnùsson, M.M., 2003, Recommendations for the prediction of avalanches, NEDIES project, 10-19, Reykjavik, Iceland, Department of Research and Processing, Icelandic Meteorological Office Nairz, P., Kriz, K., Kinberger, M., 2004, ONLINE DECISION SUPPORT TOOL FOR AVALANCHE RISK MANAGEMENT, Proceedings of the 2004 International Snow Science Workshop, Jackson Hole, Wyoming, USA OGC, Open Geospatial Consortium, http://www.opengeospatial.org


ADVANCES IN AVALANCHE FORECASTING, PODBANSKÉ


RUDI MAIR, PATRICK NAIRZ

ADVANCES IN AVALANCHE FORECASTING, PODBANSKÉ AVALANCHE W ARNING SERVICE TYROL, INNSBRUCK

AVALANCHE DANGER PATTERNS A new approach to snow and avalanche science Rudi Mair*, Patrick Nairz Avalanche Warning Service Tyrol, Innsbruck, Austria INTRODUCTION We encounter patterns constantly in everyday life. As applicable to the avalanche danger patterns introduced here, the challenge is to recognize clearly delineated, frequently recurring, blatently perilous situations over the course of highly varied winters. The fact is, accident analysis over the last twenty years has made it abundantly evident that two small handfuls of danger patterns have been responsible for the overwhelming majority of avalanche accidents. Assuming adequate knowledge and appropriate conduct, most of these avalanche accidents would be avoidable. DANGER PATTERNS - A COMBINATION OF WEATHER, SNOW LAYERING AND AVALANCHE ACTIVITY Of preeminent importance is to recognize which combination of snow layering (itself the consequence of just previous weather conditions) plus weather conditions immediately following, currently prevails. That is what brings about, i.e. determines the character and extent of the unfolding avalanche. Backcountry skiers with sufficient experience can manage to call forth (often unconsciously) stored knowledge of such avalanche scenarios from their rich memory trove of experiences and to adapt their conduct accordingly. However, most backcountry skiers simply lack the time for an intensive study of snow, weather conditions and avalanches. For that reason, they simply perceive a snowcovered landscape, without having an inkling of the dangers lurking behind it. AN OVERVIEW OF THE INDIVIDUAL DANGER PATTERNS Initially arising out of a vague notion of composing a list of loosely defined individual danger patterns, an utterly cohesive system has crystallized over time. In the course of its development, the 10 decisive danger patterns were ascertained and defined which cover no less than 98% (at least) of all danger situations which occur during the course of a given winter. The hallmark of these patterns lies in __________________________ * Corresponding author address:

Rudi Mair Avalanche Warning Service Tyrol Herrengasse 1-3 6020 Innsbruck, AUSTRIA e-mail: rudi.mair@tirol.gv.at

their repeated recurrence, not merely (although mostly) over the course of a single winter season, but above all, over a series of different, highly varying winters. The 10 most significant avalanche danger patterns at a glance: dp.1 the second snowfall dp.2 full depth snowslide dp.3 rain dp.4 cold following warm / warm following cold dp.5 snowfall after a long period of cold dp.6 cold, loosely packed new fallen snow plus wind dp.7 shallow snow areas in a season of heavy snowfall dp.8 surface hoar blanketed with fresh fallen snow dp.9 graupel blanketed with fresh fallen snow dp.10 springtime scenario DANGER PATTERN SUBDIVISIONS The individual danger patterns can be subdivided in terms of seasonality, physical area and according to their so-called ‗threat‘. For example, the pattern ‗the second snowfall‘ occurs particularly frequently in November and December, whereas the pattern ‗springtime scenario‘ is typical of the months March and April. Broken down in terms of physical area, the individual danger patterns can be distinguished by their altitude, exposition and region. The subdivision ‗threat‘ denotes the urgency and potential extent of danger inherent in a given pattern. For example, the pattern ‗snowfall after a long period of cold‘ is responsible for about 25% of all avalanche accidents, whereas the pattern ‗graupel‘ numbers among the rarely occurring patterns, responsible for only a small proportion of accidents (but, is for that very reason a ‗trap for experts‘). ACCIDENT EXAMPLES AND BACKGROUND KNOWLEDGE Of particular significance for accident prevention is a clear and easily graspable depiction of individual danger patterns. To that end, a high-impact method combining an ‘actual


RUDI MAIR, PATRICK NAIRZ

ADVANCES IN AVALANCHE FORECASTING, PODBANSKÉ AVALANCHE W ARNING SERVICE TYROL, INNSBRUCK

FIGURE 1. ‘Threat’ of given danger patterns (dp)

FIGURE 2. Physical area/seasonality depiction of dp.4

FIGURE 3. Avalanche accident Metzen 03.01.2010 (gm.4)

avalanche accident‘ and explanatory ‗background knowledge‘ was chosen. In other words, de facto avalanche accidents were described and analysed, assigned to the appropriate danger pattern and finally, rounded out with the necessary details of meteorology, snow and avalanche science, accompanied by corresponding graphs. REFERENCES Mair R., Nairz P., (2010). lawine. die 10 entscheidenden gefahrenmuster erkennen. Tyrolia Verlag Innsbruck, ISBN 978-3-7022-3086-9.

FIGURE 4. Snow layering ‗cold following warm‘ (dp.4)


Backcountry Risk Analysis for Fracture Depth and Slope Angle D.M McClung* Department of Geography University of British Columbia, Canada EXTENDED ABSTRACT In North America and Europe about 90% of deaths due to snow avalanches are triggered by backcountry travelers. Avalanche forecasts for the backcountry are from warning centers and they are freely available. Avalanche forecasts from centers are typically made on a synoptic scale but backcountry travelers must make decisions on a much smaller scale. In this paper, very simple risk analysis is given to help those decisions based on what can be measured prior to the decision. Two measureable quantities about dry slab avalanche release prior to release are the depth to the weak layer and the slope angle. Both are important in risk analysis. As the slope angle increases, the probability of avalanche release increases dramatically. As the slab depth increases, the consequences increase if an avalanche releases. In this paper, three data sets compiled from slab avalanche fracture lines are analyzed from the perspective of probability of release and risk from measured slope angles and fracture depths. It is concluded that fracture depths follow a generalized extreme value (GEV) probability density function (pdf) and slope angles tend to follow log-logistic type pdfs. The data sets are sorted by trigger mechanism: (1) natural triggers such as snowfall ; (2) a mix of triggers (natural, human triggered and explosive control); and (3) skier triggering. It is concluded that the data skewness is needed differentiate between the data sets by triggering mechanism. Special attention is given to risk for skier triggering due to the large number of deaths from slab avalanches in Europe and North America. Risk is approached from the product of the probability of skier triggering and consequences of triggering. __________________________ * Corresponding author address:

D. M. McClung Department of Geography University of British Columbia Vancouver, B.C., CANADA tel: 604-822-3537; e-mail: mcclung@geog.ubc.ca

Simple risk analyses for skier triggering are included for both slope angle and slab depth. The analysis suggests that there is a range of slab depths about 0.6 -1.0 m for which the risk of death is highest. For slope angles, there is also a range between about 33° - 45° for which the risk is highest. For the slab depth analysis, it is shown that the consequence portion for the risk of death increases by an order magnitude as the avalanche size (Canadian scale) increases by one size. REFERENCES Bair, E.H., Dozier, J. and K.W. Birkeland (2008) Ava lanche cro wn -de pth distributions, 23502,doi:10.1029/2008GL035788, 2008. Baţant ZP, Zi G and McClung D (2003) Size effect law and fracture mechanics of the triggering of dry snow slab avalanches. doi:10.1029/2002JB001884. Benjamin, J.R and C.A. Cornell (1970) Probability, statistics and decision for civil engineers, McGraw-Hill Inc., New York, 684 pp. Bezzola, JR (2012) Personal communication. CAA (2011) Canadian Avalanche Association, http://www.avalanche.ca/cac/library/incident report-database/view. Christensen, ED and Lacsina EQ (1999) Mountaineering fatalities on Mt. Rainier, Washington, 1977 – 1997: autopsy and investigative findings. Am . J. Foren. Med. Path. 20: 173 -179. Dowling, Claudia Glenn (1996) Death on the mountain. Life , August,1996: 42. Fung YC (1965) Foundations of solid mechanics, Prentice-Hall Inc. Englewood Cliffs, NJ U.S.A. 525 pp. Grimsdóttir H (2004) Avalanche risk m a n a g em e n t in b a ckc o u n t r y sk i i ng operations. M.Sc. Thesis, University of British Columbia, 173 pp. Grimsdóttir H and McClung D (2006) Avalanche risk during backcountry skiing – an analysis of factors. Nat. Hazards 39: 127-153. Jamieson B and Geldsetzer T (1996) Avalanche accidents in Canada, Vol. 4: 1984 – 1996. Canadian Avalanche Association, Revelstoke, B.C., Canada, 193 pp. McClung DM (1979) Shear fracture precipitated by strain softening as a mechanism of dry slab avalanche release. J. Geophys. Res. 84 (B7) : 3519 – 3526.


D. M. MCCLUNG DEPARTMENT OF GEOGRAPHY, UNIVERSITY OF BRITISH COLUMBIA

McClung DM (1980) Creep and glide processes in mountain snowpacks, National Hydrology Research Institute, paper no. 6, Inland Waters Directorate, Environment Canada, 66 pp. McClung DM (1981) Fracture mechanical models of dry slab avalanche release. J. Geophys. Res. 86 (B11): 10783 – 10790. McClung DM (2008) Snow avalanches as a noncritical punctuated equilibrium system. Nonlinear Dynamics in Geosciences , A.A. Tsonis and J.B. Elsner (Eds), Springer, New York, 429 – 456. McClung DM (2009) Dimensions of dry snow slab avalanches from field measurements. J. GeophysRes.114:F01006.doi:10.1029/2007JF00 0941. McClung DM ( 2011) The strength and weight of evidence in backcountry avalanche forecasting. Nat. Hazards 59: 1635 – 1645; doi 10.1007/s11069-011-9856-y McClung DM and Borstad CP (2012) Deformation and energy of dry snow slabs prior to fracture propagation. J. Glaciol. 58(209) doi:10.3189/2010JoG11J009. McClung D and Schaerer P (2006) The avalanche handbook (3 rd Edition), The Mountaineers Books, Seattle, WA., 342 pp. McClung DM and Schweizer J (1999) Skier triggering, snow temperatures and the stability index for dry slab avalanche initiation. J .Glaciol . 45(150): 190 – 200. McClung DM and Schweizer J (2006) Fracture toughness of dry snow slab avalanches from field measurements. J. Geophys. Res.111,F04008,doi:10.1029/2005JF000403,200 6. Reiweger I and Schweizer J (2010) Failure of a layer of buried surface hoar. Geophys. Res. Let t. ,37 , L24501, doi:10:1029/2010GL045433. Schweizer, J and Camponovo C (2001) The skier‘s zone of influence in triggering slab avalanches, Ann. Glaciol. 32: 314 – 320. Schweizer J and Jamieson JB ( 2001) Snow cover properties for skier triggering of avalanches. Cold Reg. Sci. and Tech. 33: 207 – 221. Schweizer J , Jamieson B and Schneebeli M (2003) Snow slab avalanche formation. Rev. of Geophys. 41(4) : 1016. Van Herwijnen, A. and J. Heierli (2009) Measurement of crack-face friction in collapsed weak snow layers, Geophys. Res. Lett 6,L23502,doi:10.1029/2009GL040389, 2009. Vick, S.G (2002) Degrees of belief. ASCE Press, Reston, VA,USA, 455 pp. Windsor JS, Firth PG, Grocott MP, Rodway GW and Montgomery HE (2009) Mountain mortality: a review of deaths that occur during recreational activities in the mountains.Postgraduate Medical Journal 85: 316 -321, doi:10.1136/pgmj.2009.0788824.

ADVANCES IN AVALANCHE FORECASTING, PODBANSKÉ


M. VALT, I. CHIAMBRETTI, P. DELLAVEDOVA AINEVA, TRENTO FONDAZIONE MONTAGNA SICURA, COURMAYEUR

ADVANCES IN AVALANCHE FORECASTING, PODBANSKÉ

YETI - a tool at the service of avalanche forecasters M. Valt,1 I. Chiambretti,1,*, P. Dellavedova2 1 AINEVA, Trento (TN), Italy 2 Fondazione Montagna Sicura, Courmayeur (AO), Italy ABSTRACT AINEVA‘s Regional Avalanche Forecasting Services uses unique software for collection, management, validation and storage of snow data. The software, developed by Barberis srl, operates under Microsoft Windows and is supported by Microsoft Access d-base. The first release was created in 1995 at the same time as in Italy had been introduced the new classification of seasonal snow. Afterwards, software had been updated developing new implementations in compliance with the needs of the Forecasting Services. In 2009, AINEVA developed the new version called YetiNik which includes IACS 1990 and 2009 Classification of seasonal snow on the ground. Currently an online version of the software is under the design stage (Husky) to allow for greater usability and to reduce maintenance costs. Yeti is written in XML language and is ready for import/export in CAAML according to international standards approved by EWAS. YetiNik is a modular software and includes several utilities: YetiMobile – transmits measurement data from mobile phone, YetiBZ – allows to enter daily survey data by telephone keypad, YetiStat – allows statistical analysis of daily survey data, YetiMap – displays spatially detected data and first level analysis. The Yeti software also has a number of utilities for sending/receiving data by FTP protocols in different formats (xlm, jpg, pdf) as well as a multilingual dictionary, built in tabular format, which can be modified by the user. Each avalanche forecasting service can use its national language or the native language for data management.YetiNik is made up by an installation package and an executable file (4 MB in size). __________________________ * Corresponding author address:

Igor Chiambretti Associazione Interregionale Neve e Valanghe - AINEVA, Trento (TN), ITALY tel: +39 0461230305 fax: +39 0461232225 e-mail: igor.chiambretti@aineva.it

Each updates provide only the exe file. The latest version is actually the 5.10. Yeti‘s development is controlled by AINEVA‘s avalanche forecasters group and its various utilities are implemented targeting the operational and management forecaster‘s needs. Software can process data to: evaluate Lemons or new Lemons for each recorded strata (it is also possible to change the threshold values); store and display the results of stability tests (RB, ECT, PST, CT); automatical-ly convert them between the IACS 1990 or 2009 classification; calculate temperature gradient for each strata, etc.. evaluate automatically, using YetiMap, snow profile‘s stability (weak, moderate, good) based on an AINEVA‘s ram profile classification (a matrix composed of 17 ram profile‘s types with different thicknesses of snow on the ground). Data spatial visualization (YetiMap) is highly flexible according to the avalanche forecaster‘s specific needs. The software distribution, in compliance with AINEVA‘s policy, is free of charge to all avalanche forecasting services financially contributing to Yeti‘s further development. INTRODUCTION In Italy, since 1984, AINEVA (Snow and Avalanche Interregional Association) worked for the standardization of data collection relating to snow and avalanches onto the Italian Alpine Arc. The Association, at first, took example from the realities existing in neighboring France and Switzerland, adopting methodologies and coding already in use. Initially, the eight Alpine regions, associated to AINEVA, had independently developed and used software for the collection, processing and management of data related to daily monitoring (AINEVA model 1) and those related to the snowpack‘s stratigraphy and weekly ram profile (AINEVA models 2, 3 and 4). In the mid-90s, AINEVA developed the idea of creating a single application for collection, storage, analysis and data management of snow science, data for avalanche forecast and research: YETI.


M. VALT, I. CHIAMBRETTI, P. DELLAVEDOVA AINEVA, TRENTO FONDAZIONE MONTAGNA SICURA, COURMAYEUR

During such software development, the AINEVA‘s regional avalanche forecasters have also decided to change some snow‘s parameters and methods of observation. These changes were designed to facilitate avalanche forecaster‘s work. For example, Wind on the station parameter, in the daily monitoring, has been abandoned (Cagnati, 2003) in favour of the observation of high elevation wind presence and activity of erosion, transportation and deposition; other changes were made to the type of observed avalanches, introducing a classification closer to that included in the avalanche danger scale (Cagnati, 1993). In the early-00s, AINEVA introduced a new model of information collection onto snow distribution on the ground, snow‘s surface characteristics, on the snow line and snow level and observation of avalanche‘s natural or artificial release. This model, called AINEVA model 6 (used mainly for the itinerant surveys), has been implemented in Yeti. While working with the same software, each region has retained data‘s ownership and its own database but ensures data sharing with all other avalanche forecasting offices partner. The concept of database‘s unification and sharing, developed in 1995, is now of vital importance for avalanche forecasters due to the low entropy of these data (LaChapelle, 1980) and the necessity of survey, budget and data sources rationalization. In 2009, AINEVA developed the new software version called YETI NIK which includes both the 1990 and the 2009 versions of the International Classification for Seasonal Snow on the Ground (UNESCO, IHP-VII, IACS). Actually, AINEVA is developing a Yeti online version (Husky) in order to allow greater usability by the regional avalanche forecasting services and surveyors. Yeti is fully compatible with XML syntax and is ready for import/export in CAAML according to inter-national standards approved by EWAS. AINEVA‘s avalanche forecasters group controls yeti‘s development and its various utilities are implemented targeting the operational and management user‘s needs and following, EAWS‘s approved, international standards thus allowing for data exchange on an international level. YETI NIK STRUCTURE Software, developed by Barberis srl, is written in C# (C Sharp), operates under Microsoft Windows (WIN95, WIN 2000, WIN XP, VISTA, WINDOWS 7) and is supported by Microsoft Access d-base. AINEVA‘s strategy is to provide a modular software easily installable, even onto

ADVANCES IN AVALANCHE FORECASTING, PODBANSKÉ

FIGURE. 1 YetiNik vrs. 5.1.0

low-end PC, and upgradeable. YetiNik is made up by an installation package (15 Mb in size) and an executable file (4 MB in size). Each update provides only the exe file. The latest version is actually the 5.10. The database is annual (hydrological year) and is divided by observation model type: MOD1, MOD23 and MOD 6. The database‘s annual span is a limit given by Microsoft Access and creates some problems in the statistical analysis of several years. Yeti Nik Yeti‘s basic module is Yeti Nick which is used to collect, import, export, view and edit data and metadata for daily, weekly and itinerant direct observation or indirect evidence (sensu La Chappelle, 1966) – (AINEVA‘s models 1, 2, 3 and 6) in the field (on a laptop) as well as in the office. Data on both snow pits, ram profile, stability tests, meteorological and snow cover observations as well as avalanche occurrences may be collected and stored into the application. Each data is collected by using custom forms (AINEVA‘s models) displayed by the application and the data collection process has been carefully designed by both the developing firm and AINEVA‘s avalanche forecasters team to be user friendly, efficient and intuitive. Several input data short cuts are provided as well as drop down list and ―point and click‖ entries (used extensively) in order to allow a more efficient, rapid and robust data entry (Fig. 2). A form level validation of data is also used throughout the input phase. The application works using S.I. - metric units and several parameters (such as temperature, grain shape classes, hardness, liquid water content) can be displayed with different scale, code style, symbols or colors. Grain shapes, for each strata, can be entered in sub-classes or not. Both lightweight (5N) or standard (10N) ram data can be processed.


A multilingual dictionary, built in tabular format and easily upgradable by the user is available. Each avalanche forecasting service can use its national language or the native language for data management (Fig. 4).

FIGURE. 2 input of snowpit data

All preferences can be customized for each user. Viewing of snow pit graphs is also supported (AINEVA‘s model 4). All data collected by an individual surveyor can be written on YetiNik, locally (each one with its own small database) and can be sent, via FTP, to a regional avalanche forecasting office‘s centralized database where they can be further reviewed and edited by an avalanche forecaster and shared with other users or leveraged for snow science research. Each regional avalanche foracasting office‘s server receives data from several desktop applications, parses, and stores the data in a central database (Fig.3).

FIGURE. 4 language selection and multilingual dictionary

The Yeti‘s core is dedicated to data collection on both snow pits, ram profile and stability tests and this application has been extensively developed over the past few years. The snow pit profiles are viewable with either the 1990 and the 2009 IACS classifications (grain shape, hardness, crystal size) – (Fig. 5).

FIGURE. 5 Snowpit and ram profile

FIGURE. 3 dataset management application

Both surveyors and forecasters use Yeti‘s same version but characterized by different privileges/authentication. Imported data, into avalanche-forecasting office‘s dataset, undergo a second treatment of automatic incongruity detection as during the input phase performed locally onto the surveyor dataset (double check). Major form level errors are therefore easily detected. YetiNik‘s files are written with XML – CAAML (Canadian Avalanche Association Markup Language) syntax, allowing easy import, in delimited format, into spreadsheets and other applications.

Snow strata can be displayed with color according to IACS‘s standard or personal preferences and for each one, or for pre-set thickness, a thermal gradient can be calculated and displayed with graduated color. Lemons (sensu Jameson 1998) are calculated and displayed according to the methodology of critical layers (Monti et al, 2012) and thresholds can be changed through a user-defined table (Fig. 6). In addition, the entire snowpack SWE is calculated by integrating density values of each layer, which was not possible to measure (Valt et al. 2009). Such integration takes place due a statistical study, carried outside of Yeti, which determined, for each geographical area of the Italian Alps, the snow density for a layer with


M. VALT, I. CHIAMBRETTI, P. DELLAVEDOVA AINEVA, TRENTO FONDAZIONE MONTAGNA SICURA, COURMAYEUR

ADVANCES IN AVALANCHE FORECASTING, PODBANSKÉ

performance of mobile phones in severe outdoor conditions. In the near future the surveyor will be able to collect data and observation and deliver them, in near real-time, to the main dataset through SMS or FTP transfer protocol.

FIGURE. 6 Lemon user-defined table

comparable liquid water content, shape of grains and hardness (Valt et al., 2012). The results of stability tests (RB, ECT, CT, PST) can be stored and displayed alongside the snow pit profile (multiple stability test for each snow pit can be recorded). Finally, a text box with further textual observations can be displayed as well. A section of the application is dedicated to the recording of data and observations made during itinerant surveys (AINEVA‘s model 6). Surveyor can record observations onto snow distribution on the ground, snow‘s surface characteristics, on the snow line and snow level and observation of avalanche‘s natural or artificial release. There is not special processing for these data. YETI NIK UTILITIES YetiNik is a modular application and each utility interact with the database in read / write or read-only modality following the settings of system‘s administrator. The YetiNik application consists of five major utilities: YetiMobile - transmission of measurement data from mobile phone; YetiBZ - entering daily survey data by telephone keypad; YetiStat - statistical analysis of daily survey data; YetiMap - displaying spatially detected data (AINEVA‘s Models 1 and 6) YetiMap_S –a first level analysis of SWE, Lemons, an automatically evaluation and classification of ram profiles. Yeti Mobile A window mobile app for daily or periodic (AINEVA‘s model 1, 2, 3 and 6) data and observation collection has been produced while an Android based one is still under de v e l opm en t d ue t o t h e g en er a ll y p o or

Yeti BZ Since 2011, surveyors without software and mobile phone coverage, however, have the ability to transmit daily data, to a dedicated phone number, typing them onto the phone keypad following a guided way. An appropriate control routine indicate, in real time, any errors and data not compatible; the receiver forecasting office uses the utility YetiBZ which connects to the phone answering machine retrieving such data. Yeti Stat This utility produces basic statistical analysis to assist the daily and long-term assessments of the avalanche forecaster. Yeti Map This utility provides data spatial visualization, onto a geographic display, up to four data or information for each observation station (upon user setting) – (Fig. 7).

FIGURE. 7 YetiMap display

Yeti Map_S This utility provides data spatial visualization and automatically evaluate snow pit and ram profile, dividing them in the following classes of snowpack stability: weakly bonded; moderately to weakly bonded; moderately well bonded; well bonded (stable). Snowpack stability is assessed using an AINEVA‘s snow pit and ram profile classification (a matrix composed of 17 profile‘s types with different thicknesses of snow on the ground). A specific stability class is assigned to each profile‘s type and thickness. For each station, the forecaster can display into a small box icon the following data:


height of the snowpack (HS); the maximum number of Lemon (L); the height from the ground (HSL) in which is positioned the surface with the maximum number of lemon; in the case of multiple layers with the same maximum number of Lemon, the shallower one is chosen. In a second displays, the forecaster can see into a small box icon, for each station, the following data: the SWE for each profile; height of the snowpack (HS). Clicking onto both box icons will display the original data in a complete stratigraphy profile. In a third displays the box icon offers a simplified ram or statigraphic profile shape. Each profile, regardless of its height, is divided into 10 parts of costant height and for each one an average hardness value is calculated. The values obtained are further classified into 10 classes of hardness. The profile is then replotted onto a grid of 10 unit‘s thickness and 10 units of hardness and HS value (height of the snowpack) is displayed (Fig. 8).

budget. New releases are tested, before diffusion, by a small group of regular users that install on several different PC and test thoroughly the beta executable file. Each new release is issued only after passing various tests. The software distribution, according to AINEVA‘s policy, is free of charge to all avalanche forecasting services financially contributing to Yeti‘s further development. CONCLUSIONS YetiNik is an application for the collection, processing, analysis, storage and management of snow science data for avalanche forecast and research and can be easily installed onto a normal commercial PC. Its versatility is due to its modular architecture and its development is conceived in compliance with surveyor‘s and avalanche forecaster‘s needs. The future development of an online version of the software will enhance its usability and reduce maintenance costs. However, the standalone version will be kept updated. REFERENCES

FIGURE 8. YetiMap snowcover stability display

Yeti Map_S utility is very practical for avalanche forecaster as, at a glance, he/she has a picture of the type of union respects. DATA OUTPUT AND IMPORT/EXPORT YetiNik can output data in classical format (JPG, PDF) or also in TIF, BMP (different image qualities can be selected). Data can also be imported/exported using the CAAML – XML syntax. The application has been designed to perform manual and automated (scheduled) procedures of sending/retrieving data by FTP protocols. S O F T W AR E DEVELOPM ENT AN D MAINTENANCE The development projects is controlled by AINEVA‘s avalanche forecasters group and its various utilities are implemented in compliance with the operational and management forecaster‘s needs., while AINEVA‘s technical board provides supervision and allocate s

Cagnati, A., 1993. La nuova scala unificata per la classificazione del pericolo da valanghe. Neve e Valanghe, V. 19, p. 26-31. Cagnati, A., 2003. Sistemi di Misura e metodi di osservazione nivometeorologici. AINEVA, Trento, Italy, 186 pp. Colbeck, S.C., Akitaya, E., Armstrong, R.I., Gubler, H., Lafeuille, J., Lied, K., McClung, D.M., and Morris, E.M., 1990. The International Classification for Seasonal Snow on the Ground. Int. Commission on Snow and Ice (IAHS), World Data Center A for Glaciology, University of Colorado, Boulder, CO, USA. Fierz, C., Armstrong, R.I., Durand, Y., Eychevers, P., Greene, E:, McClung, D.M., Nishimura, K., Satyawali, P.K., and Sokratov, S.A., 2009. The International Classification for Seasonal Snow on the Ground. IHP-VII Technical Documents in Hydrology N°.83, IACS Contribution N°.1, UNESCO-IHP, Paris, France, 80 pp. LaChapelle E.R., 1966. Avalanche forecasting – a modern synthesis. Int. Symposium on Scientific Aspects of Snow and Ice Proceedings. Int. Assoc. Sc. Hydrology, N. 69, p. 350-356. LaChapelle E.R., 1980. The Fundamental processes in Conventional Avalanche Forecasting. Jour. of Glaciology Vol. 26, No.94, p. 75-84. Jamieson, B. and Schweizer, J., 2005. Using a checklist to assess manual snow profiles. Avalanche News 72, Canad. Avalanche Assoc., Revelstoke, BC., Canada, p. 72-61. Monti, F., Cagnati, A., Valt, M. and Schweizer, J., 2012. A new method for visualizing snow stability profiles. Cold Regions Science and Technology, v. 78(o), p. 64-72.


M. VALT, I. CHIAMBRETTI, P. DELLAVEDOVA AINEVA, TRENTO FONDAZIONE MONTAGNA SICURA, COURMAYEUR

Valt, M., Cianfarra, P., Cagnati, A., Chiambretti, I. and Moro, D., 2010. Estimate of snow density knowing grain and share hardness. Geophysical Research Abstract - Vol. 12, EGU 2010-122172010. Valt, M., Monti, F., Cianfarra, P., and Moro, D., (1) 2012. Physical properties of snow cover in the Alps - insight from the Davos area (Switzerland) and Veneto-Friuli Venezia Giulia regions (Italy). Geophysical Research Abstracts, Vol. 14, EGU2012-12471, 2012, EGU General Assembly

2012.

ADVANCES IN AVALANCHE FORECASTING, PODBANSKĂ&#x2030;


MARTIN VOJTEK MILITARY W EATHER CENTRE, SLOVAK AIR FORCE

ADVANCES IN AVALANCHE FORECASTING, PODBANSKÉ

LABINIX – a tool for regional avalanche danger assessment from meteorological data based on regression Martin Vojtek* Military Weather Centre, Slovak Air Force INTRODUCTION Computer-assisted models help the avalanche forecaster to gain confidence and reduce the risk of failure caused by the human factor. LABINIX is a simple software tool that calculates daily avalanche danger level from meteorological data. The name comes from Latin: LABIna = avalanche and NIX = snow. So far, the area of interest covers the High Tatra Mts. (~ 250 km2), and the measurements from 7 meteorological stations (Figure 1) are used as input. METHOD The method based on stepwise regression is described in detail in author‘s PhD thesis (Vojtek, 2010). Beside the basic input meteorological variables, also some nonlinear, mixed and cumulative variables (up to 7 days back) were defined. Another useful element – snow water equivalent (SWE) – is measured only once a week in Slovakia. Therefore, a method developed by Němec et al. (2003) is adopted in LABINIX to estimate the daily SWE values. The dataset was divided into 4 subsets, depending on the water content of avalanches fallen (dry, wet and moist/combined) and the rest consisting of non-avalanche days. The best set of avalanche-danger-level-related variables was selected by stepwise regression using both stepping direction approach in Splus (Mathsoft, 1999). The models are signed by G (general), D (dry), M (moist) or W (wet), followed by number of stations and number of variables used. Regression coefficients were calculated on the period 1994-2002 plus some days with estimated avalanche danger levels 3 and more (selected from historical records back to 1980). RESULTS The best model D7x44 predicts 74% of days correctly on testing dataset (2003-2008). The __________________________ * Corresponding author address:

Martin Vojtek Military Weather Centre, Slovak Air Force Na Slatinkách 7 962 31 Sliač, SLOVAKIA tel: +421 949 338 609 e-mail: vojteks@gmail.com

error is not greater than 1.5 for the rest of the days. If the best model is selected for each month, the accuracy ranges between 66% in May to 82 % in December, (refer to Table 1). The main drawbacks: 1. if a meteorological station moves, the regression coefficients must be recalculated; 2. if a meteorological element is missing, the avalanche danger level cannot be calculated until the value is estimated / completed later; 3. a l l models have problem wi t h underforecasting of the avalanche danger level 3 (considerable). On the other hand, the avalanche danger level can be forecasted for the next 2-3 days if data for the following days are inserted from a numerical weather prediction model (e.g. ALADIN). LABINIX can be thus useful for skitour planning or earlier warning for extreme avalanches. In order to use the LABINIX operationally, further recalculations were performed due to lack of some meteorological elements in realtime. At the moment, LABINIX is capable to load and decode 2 special meteorological messages / bulletins (INTER and INTER TATRY, available at 06:50 UTC), to check for missing data, and to calculate the avalanche danger level (using M7x60 designed for moist avalanche days). The tool offers a forecaster variety of charts, e.g. contributions of meteorological elements to the calculated avalanche danger level (Figure 2). The best avalanche-related elements are: snow depth, air temperature, relative humidity, snow water equivalent, SDRRH (introduced by Salway, 1979), the daily amplitude of air temperature, and new snow height. CONCLUSIONS There is a growing potential for numerical avalanche forecasting in Slovakia thank to increasing records of avalanche, meteorological, snow profiles and stability data. Automatic weather stations introduced in 2007/08 will be another useful input. LABINIX is going to be used in the following winter as experimental tool, and remains the subject of further development. If it proves to be helpful, it can be applied to other avalanche regions in Slovakia.


MARTIN VOJTEK

ADVANCES IN AVALANCHE FORECASTING, PODBANSKÉ MILITARY W EATHER CENTRE, SLOVAK AIR FORCE

FIGURES, TABLES, AND PHOTOS

REFERENCES

TABLE 1. The proportion of correct PC= (H1+H2+H3+H4+H5)/N, where H1,...,H5 are correctly forecasted avalanche danger levels, i.e. the difference between the forecasted and the real value is less than 0.5. N is the number of all days in whole period (all) or in months (December to May).

Model

all

D

J

F

M

A

M

G7x19

72

81

81

75

81

54

66

D7x44

74

82

78

76

68

64

60

M7x54

69

82

70

67

73

51

47

W7x17

71

82

72

69

67

58

53

6 5 4

SD, T, RH, SWE, SDRRH, Tamp, NS, other

Mathsoft, Inc. (1999): S-PLUS 2000 Guide to Statistics, Vol.1. Data Analysis Products Division. MathSoft, Inc., Seattle Němec, L., Květoņ, V., Setničková, I., Škáchová, H., 2003. Estimation of the water equivalent of snow cover from the other meteorological measurements. ICAM-MAP Meeting 2003, Brig, Switzerland, (Publications of MeteoSwiss, No.66.) Salway, A.A., 1979. Time-series modelling of avalanche activity from meteorological data. Journal of Glaciology, Vol. 22, No. 88, p. 513528. Vojtek, M., 2010. The dynamics of snow cover in mountainous regions of Slovakia. Ph.D. Thesis, Comenius University, Bratislava, Slovakia, 119 pp.

FIGURE 1. Slovak Hydrometeorological institute meteorological stations: Podbanské (972 m), Štrbské Pleso (1322 m), Ţdiar-Javorina (1013 m), Lomnický štít (2635 m), Skalnaté pleso (1778 m), Tatranská Lomnica (827 m). Chopok (2005 m) is in Low Tatra Mts.

3 2 1 0 -1 -2

FIGURE 2. Contributions of the most significant variables to the value of forecasted avalanche danger level (red line): SD = snow depth, T = air temperature, RH = relative humidity, SWE = snow water equivalent, SDRRH = multiplied SD with rainfall with RH, Tamp = air temperature amplitude, NS = new snow height; and other variables. Dots are forecasts issued by APC. Example from 15th Dec 2005 to 5th Jan 2006


ADVANCES IN AVALANCHE FORECASTING, PODBANSKÉ


ADVANCES IN AVALANCHE FORECASTING, PODBANSKÉ


PAVOL BERÁNEK, MARTINA SADLOŅOVÁ DEPARTMENT OF NATURE PROTECTION THE KRKONOSE MTS. NATIONAL PARK

ADVANCES IN AVALANCHE FORECASTING, PODBANSKÉ

Analysis of weather condition on 25th March 2009 over Slovakia and numerical weather prediction outputs for the day of avalanche occurrence in Žiarska valley Pavol Beránek, Martina Sadloňová* Slovak Hydrometeorological Institute, Bratislava, Slovakia ABSTRACT In NW cyclonal situation during last few days before 25th March there was heavy snowfall over interest area. In combination with strengthening wind and temperature drop below -8°C it triggered enormous avalanche. From numerical prediction outputs (both Aladin and ECMWF models) it was a few days before event clear that for the day of avalanche occurrence temperature would drop a lot and a big amounts of new snow would fall. The closer to event date the more accurate forecast of both models were. KEYWORDS Synoptic situation, circulation, precipitation, snow depth.

__________________________ * Corresponding author address:

Martina Sadloňová Slovak Hydrometeorological Institute Jeséniova 17 833 15 Bratislava, SLOVAKIA tel: +421 259415370 e-mail: ovp@shmu.sk


ANNA SERES INSTITUTE OF GEOGRAPHY UNIVERSITY OF MISKOLC

ADVANCES IN AVALANCHE FORECASTING, PODBANSKÉ

Weather, terrain, vegetation and snowpack based avalanche risk model Anna Seres* Institute of Geography, University of Miskolc, Hungary INTRODUCTION Avalanche accidents become more and more frequent with the increasing number of people going for winter outdoor activities to the mountains. Daily updated danger maps are available for most of the mountains which get a lot of tourism in the winter. However these maps only indicate a general level of avalanche danger, sometimes giving some more details on the height and aspect where the danger is higher. These maps usually work as a go/not go to that specific mountain decision aid and they are not sufficient for proper and safe route planning. Route planning has to be done on site, which requires experience and knowledge of the area. AIMS My aim is to prepare a model which creates daily updated, very detailed, good resolution avalanche risk maps, showing the actual risk of the snowpack in each valley, slope, ridge, etc, giving more details to the present danger maps, thus making it easier for the general public to avoid the dangerous areas. STUDY SITE The area of my study is the middle part of the Low-Tatras in Slovakia, around the peak Chopok. The area has widespread tourism with several ski slopes, a huge development project is being carried out with new lifts at the moment, and perfect slopes for ski touring. MATERIALS Avalanche danger basically depends on four factors: weather, terrain, snowpack and vegetation cover/surface type. The model is developed based on data from the winter of 2009/2010. Weather and snowpack data: minimum and maximum temperature, wind speed, wind direction, cloudiness, humidity, new snow height, new snow density (from snow height and water equivalent), snowpack height and snow crystal type was provided by __________________________ * Corresponding author address:

Anna Seres Institute of Geography, University of Miskolc Erzsebet setany 69 3517 Miskolc-Lillafured, HUNGARY tel: +36-20-3452263 e-mail: anna.seres@gmail.com

SLP (Avalanche Prevention Centre of Slovak Mountain Rescue Service) and SHMÚ (Slovak Hydrometeorological Institute). Weather and snowpack data was available for the maximum of 6 meteorological stations on the area, most type of data only for 2-4 stations. DEM was created by digitizing the contour lines of 1:10000 topographic maps with resolution of 10m. Vegetation was defined by supervised classification of Landsat images. Results are validated by snow profile data, taken approximately weekly from the area. METHODS The snowpack model follows the development of the snowpack from the first snowy day to the spring melt. The development of the snowpack is estimated based on weather and terrain data. Minimum, maximum temperature and wind is interpolated from point measurements to the whole area. In the model each new snowfall exceeding 5 cm, each day with strong wind, depositing significant amount of snow or each day with surface hoar or ice formation results a new layer in the snowpack. Each layer has new snow height, new snow density and snow crystal type as input starting attributes. These properties change from day to day according to the input weather and terrain parameters and snow crystal size and layer stability is added. The model creates weak layers in the snowpack at locations depending on weather and terrain, but if these weak spots really mean avalanche danger or not is decided by the model by further evaluation based on terrain parameters and vegetation cover. The model‘s input requirements are: weather and new snow data, DEM derived terrain parameters and vegetation cover/surface type from remotely sensed data. It creates intermediate weather and snowpack maps, interpolating from measures point data. The snow stability map, showing the weak layers, together with the potentially dangerous areas located by terrain and vegetation cover, at the end create a daily updated, detailed avalanche risk map, which can be used by itself or can make the existing, general danger level more detailed. (Figure 1.)


ANNA SERES INSTITUTE OF GEOGRAPHY UNIVERSITY OF MISKOLC

ADVANCES IN AVALANCHE FORECASTING, PODBANSKÉ

FIGURE 1. Flowchart of the avalanche risk model. Paralelograms show constants, flags show maps and rectangles show processes. White flags provide explanation.


ANNA SERES

ADVANCES IN AVALANCHE FORECASTING, PODBANSKÉ

INSTITUTE OF GEOGRAPHY UNIVERSITY OF MISKOLC

RESULTS Weather module Temperature Temperature maps will be used as input for snowpack compaction, crystal metamorphism and snow stability equations. Minimum and maximum temperature was calculated for each day using multiple linear regression. The explanatory variables are DEM and solar radiation (SR), this latter calculated based on the cloudiness in the morning of the actual day (SRact). The dependent variable is the minimum/maximum temperature for the 5 weather stations. Solar radiation was calculated with the solar radiation tool of ArcGIS for each day, taking the azimuth, altitude, proportion of disuse radiation and transmissivity of air into account. Regression was also solved for variable combinations: DEM + SR8(cloudy) and DEM + slope + aspect, this latter calculated for each day, taking the azimuth and altitude of the warmest time of day into account. Pearson‘s correlation and statistical tests showed better results for the DEM + slope + aspect regression, than for the DEM + SR regressions (Table 1-4.), but these numbers cannot be trusted as there are only 5 points to calculate from. TABLE 1-2. Values of statistical test for different variable combinations(1.maximum, 2. minimum temperature). Cells in green mark the best results. maximum temperature

DEM_SR8

DEM_SRact

DEM_aspect_ slope

RMSE Adjusted R2 T-Prob F-Prob

1,029 0,590 0,427 0,205

0,965 0,625 0,357 0,188

0,526 0,675 -1,000 0,275

DEM_SR8

DEM_SRact

1,403 0,264 0,486 0,325

1,417 0,248 0,463 0,337

minimum temperature RMSE Adjusted R2 T-Prob F-Prob

DEM_aspect_ slope 1,163 0,213 0,448 0,410

TABLE 3-4. Pearson’s correlation vales for different variables maximum temperature Pearson' correlation minimum temperature Pearson' correlation

DEM

SR8

Sract

aspect

slope

0,82 0

0,73 0

0,57 3

0,491

0,29 5

DEM

SR8

Sract

aspect

slope

0,65 1

0,56 9

0,41 2

0,331

0,25 8

By looking at the map and checking its maximum, minimum values and the spatial

distribution of the temperature, the DEM + SRact combination seemed to be the best version (Figure 2.). The DEM + slope + aspect also showed good maximum and minimum values, but the spatial distribution was not correct, as it does not show the shadowing effect of the surrounding terrain. Wind Wind can redeposit huge amounts of snow, adding an extra load to the snowpack and potentially trigger avalanches. Wind is measured in every 10 minutes on 2 automatic weather stations, Jasna and Chopok. Only wind speeds above 10 m/s are considered, as that is the approximate threshold of drifting snow (McClung et al. 1999.). These are then averaged for different directions for each day. These wind speeds are interpolated on macroscale by the empirical calculation of Hellman exponent in the Hellman equation by Spera et al. 1979 and modified on a micro scale by the slope, aspect and curvature based formulas of Liston et al. (2007). Areas with sudden change in wind speed are the most prone to gather snow, so these were selected by calculating a range of 5 cells radius on wind speed. Wind drift is considered only at locations with grass (or rocks) or dwarf pine vegetation. So the wind speed map is masked with the vegetation (Figure 2.). Snowpack module The snowpack module of the model has not been finalized yet, results are expected later this the year. New snow height, density Newly fallen snow gets denser and looses height with time. This process is the fastest in the first few days and it depends on temperature and later, the weight of the snow above. Snow compaction is calculated by the formula of Flerchinger at al. (1989) and McConkey (1992). Snow crystal type and size Metamorphism of the snow mainly depends on the temperature gradient in the snowpack, which is the function of snowpack height and surface temperature. Snowpack height is calculated as the sum of height of snow layers, bottom temperature is considered 0°C, snow surface temperature is taken the same as air tem perature. e qu i - te m perature (ET ) metamorphism takes place when the temperature gradient of the snow layer is less than 10°C/m. During ET metamorphism, the b r a n c h e s , c o r n e r s o f t h e s n o w c r ys t a l


ANNA SERES INSTITUTE OF GEOGRAPHY UNIVERSITY OF MISKOLC

ADVANCES IN AVALANCHE FORECASTING, PODBANSKĂ&#x2030;

FIGURE 2. Maximum temperature map by regression with DEM and solar radiation, based on the cloudiness in the morning of the actual day

FIGURE 3. Possible wind loading areas for wind speed Chopok 15 m/s, south (magenta shows areas with highest wind loading)


ANNA SERES INSTITUTE OF GEOGRAPHY UNIVERSITY OF MISKOLC

disappear and rounded grains bonded by small necks develop. This snow type is not likely to collapse and cause avalanche. Temperature gradient (TG) metamorphism takes place at temperature gradients above 10°C/m. It produces angular, later hollow, cup-like big crystals, which easily collapse under the weight of above snow layers and produce avalanches. Melt-freeze (MF metamorphism occurs when the temperature in the day is high enough for the snow to melt and cold enough in the night to freeze. In the day it produces weak, unstable layer, while it becomes strong and stable in the night. Physical process based equations describing the exact way and extent of crystal growth require input parameters that can only be measured in laboratories, so it cannot be applied to my model. Instead, the simple, empirical formula of Brun et al, estimating the dendricity and sphericity of the snow crystals, used in the Swiss SNOWPACK model (Lehning et al. 2002) will be used in my model, modified with empirical relations from snowpack data of the area. Surface hoar and ice layers will be generated on the surface at appropriate weather conditions. For surface hoar it means relative humidity above 70%, wind speed of 1-2 m/s, surface temperature of 0--4°C or -12--21°C, clear sky in the night for significant outgoing radiation, northerly aspect, convex forms (McClung 1999). Ice layer is formed in case of persisting melt-freeze changes or rain on snow. Snow stability map, weak layers Each snow layer has the following attributes: snow crystal type (dendricity, sphericity), snow crystal size, ice (yes/no). These attributes are calculated for all new snow, old snow layers, wind deposited layers, surface hoar and ice layers. The attributes are updated daily according to weather and terrain data. The stability of the snow layer is defined based on these attributes. Weak layers, low stability values are defined in case of large temperature gradient crystals, very wet snow, buried ice layer, buried graupel layer, buried surface hoar layer (depending on the weight of the snowpack above). Stability of weak layers also depend on temperature and aspect, as for example a buried ice layer on a southerly slope is more dangerous if the temperature is high and melting occurs, than when temperature is low. Terrain module The result of the terrain module at the end modifies the snow stability layer according to how good the terrain is for releasing

ADVANCES IN AVALANCHE FORECASTING, PODBANSKÉ

avalanches. Potential avalanche release areas were defined by slope and curvature. Slope Typical slope angles for avalanche starting zones are between 25-60°. In the most cases this is true for the Low-Tatras as well. The maximum slope angle on the study site is 58°. Biggest avalanches can form between slope angles 30-50°. Snow slides easier as the slope becomes steeper, so I considered a linear relationship between slope angles and avalanche danger on the area. As there were only few pixels of high values, I stretched the raster between 3 standard deviations to get the middle 99,7% of the data. The slope raster then was normalized between 1 and 10. Profile curvature Avalanches form easier on areas where the snowpack has no support from below, so around inflection lines, where the profile curvature changes from the convex of the ridge to the concave of the valley (Ciolli et al. 2000). The avalanche prone northern slopes of the Low-Tatras fall suddenly behind the main ridge and the avalanche starting zones in the southern side are found in valley heads. These cases rarely coincide with the inflexion line, so profile curvature was not included in the calculations. Plan curvature Avalanches in the southern slopes of the LowTatras usually have confined path and the starting zones are usually found in valley heads. These areas can be located with plan curvature. Concave forms, marked with negative values are the most likely to gather snow and slide off on the curvature raster, so first I calculated the inverse of the scale and then shifted the beginning to zero. Then it was stretched between 3 standard deviations from the mean and normalized between 1 and 10. This way the slope and curvature could be compared. Avalanche terrain factor Slope and plan curvature effect the location of the starting zones in about equal degree. So I multiplied the two before mentioned maps to get the final possible starting zones map. The result shows the unconfined avalanche paths of the northern steep slopes confined avalanche path in the valley heads of the south. Avalanche release areas were only considered at grass, rock or dwarf pine covered areas, so the result was masked with landuse/vegetation. The histogram shows


ANNA SERES INSTITUTE OF GEOGRAPHY UNIVERSITY OF MISKOLC

higher values under the starting zones, than on the whole area. Two bigger avalanche starting zones seems not to be located correctly at first sight. If looking at them closer, one can see, that the release areas in these cases are probably not the top of the paths, but the east side in the first case (1) and the bottom in the second case (2). In both cases the slide started here, but after the snow slid off from these areas, the snowpack above lost its support and the whole above lying snowpack slid off. Vegetation module Vegetation/land use is important because rocks, trees anchor the snowpack and do not let it slide. Even a very shallow snowpack can slide off of areas covered with grass, herbaceous vegetation or fine scree. Snow usually only slides off of areas with dwarf pines, after the pines are buried at snowpacks about 1 meter deep. Mature forests rarely make it possible for avalanches to start. Threshold snowpack heights can be given for areas with different vegetation to decide if avalanches can occur there or not. Land use maps were created by supervised classification of a combined Landsat and ASTER image. Landsat images were taken from a spring and an autumn date to better show the different reflectance of different vegetation types. NDVI was calculated for both dates and all layers

ADVANCES IN AVALANCHE FORECASTING, PODBANSKĂ&#x2030;

FIGURE 4. Avalanche terrain factor. Darker colors show higher risk of potential snow slide. Red: avalanche path according to the avalanche cadastre. Upper left corner: histogram of terrain factor for the whole area, upper right: histogram for the starting zones. Numbers show two avalanche paths with release areas not at the top of the path.

were stacked. The 30m resolution (same as Landsat) ASTER DEM was downloaded for the same area, the whole Low-Tatras, slope and aspect was calculated and the DEM, slope and aspect were also layer stacked behind the Landsat layers. This way when delineating the training areas, I not only show the typical reflectance values, but also the typican elevation, slope and aspect, which makes the classification more precise. Typical areas for 18 classes were delineated as training areas. Classes were merged into 3 final classes based on object height. The 3 final classes are: grass / fine scree / plough land; deciduous / coniferous forests; dwarf pines (Figure 4.). Threshold snowpack height values for these classes are: 0,2m for grass/scree/plough land, 1m for dwarf pine and snow is not allowed to slide in mature forests. Avalanche risk map The avalanche risk map is the final outcome of the model. The snow stability / weak layers, the possible avalanche release areas, the vegetation cover are combined to produce the


ANNA SERES INSTITUTE OF GEOGRAPHY UNIVERSITY OF MISKOLC

final avalanche risk map. It can be used by itself or to make the existing avalanche forecast more detailed. The risk map is validated by snow profile data taken about once a week of the area. CONCLUSION The snow stability, potential release areas, vegetation cover based risk map would be updated daily to follow the changes in the snowpack. The up to date information of the snowpack allows the forecasts to be more precise, to show the detailed spatial distribution of the avalanche risk. The snowpack module is not ready yet, but I believe that when the whole model is ready, validated and finalized, it will be a great addition to any avalanche forecasts, it would ease the route selection even for not experienced skiers and hopefully help to avoid accidents. ACKNOWLEDGEMENTS I would like to thank to the staff of SLP (Avalanche Prevention Centre of Slovak Mountain Rescue Service) for their support and help in providing weather and snowpack data and discussing the results. I would also like to thank to SHMÚ (Slovak Hydrometeorological Institute) for providing the data from the Chopok AWS. The described work was carried out as part of the TÁMOP-4.2.2/B-10/1-2010-0008

ADVANCES IN AVALANCHE FORECASTING, PODBANSKÉ

FIGURE 5. Vegetation map of the study area.

project in the framework of the New Hungarian Development Plan. The realization of this project is supported by the European Union, co -financed by the European Social Fund. REFERENCES Ciolli, M., Zatelli, P. 2000. Avalanche risk management using GRASS GIS, Geomatics Workbooks, Vol. 2000, No. 1, pp. 1-12 Lehning, M., Bartelt, P., Brown, B., Fierz, C., Satyawali, P. 2002. A physical SNOWPACK model for the Swiss avalanche warning – Part II. Snow microstructure, Cold Regions Science and Technology, Vol. 35, pp.147-167 Liston, G.E., Haehnel, R.B., Strum, M., Hiemstra, C.A., Berezovskaya, S., Tabler, R.D. 2007. Instruments and Methods – Simulating complex snow distribution in windy environments using SnowTran-3D, Journal of Glacilology, Vol. 53, No. 181, pp. 241-256 McClung, D., Schaerer, P. 1999. The Avalanche Handbook, fifth printing, Seattle, Washington: The Mountaineers, USA Spera, D.A., Richards, T.R. 1979. Modified power law equations for vertical wind profiles, Wind Characteristics and Wind Energy Siting Conference, Portland, Oregon, June 19-21. 1979. USA 12p.


R. SOLBERG, R. FRAUENFELDER S.Ø. LARSEN, A.-B. SALBERG NORWEGIAN COMPUTING CENTER (NR) NORWEGIAN GEOTECHNICAL INSTITUTE (NGI)

ADVANCES IN AVALANCHE FORECASTING, PODBANSKÉ

Experiments with remote sensing in the context of avalanche warning and detection R. Solberg1,* R. Frauenfelder2 S.Ø. Larsen1 A.-B. Salberg1 1 Norwegian Computing Center (NR), Oslo, Norway 2 Norwegian Geotechnical Institute (NGI), Oslo, Norway INTRODUCTION In many mountain regions of Norway snow avalanches pose a risk to road and railway passengers as well as tourists, skiers and others during the winter season. Each year, snow avalanches hit populated areas and parts of the transport network, leading to the damaging of buildings and infrastructure, sometimes also to the loss of lives. Much of the country is remote and knowing exactly where avalanches are likely to take place or have taken place is a challenge for the authorities. Earth observation satellites, therefore, represent a potentially important source of information. Two Norwegian projects carried out by NGI and NR have investigated and experimented with the potential of using remote sensing for avalanche warning and detection: The Norwegian Space Centre (NSC) supported project ―Improved Avalanche Warning Using Satellite Data‖ (2008-2010) and the European Space Agency (ESA) funded project ―Avalanche Inventory for Decision Support and Hind-cast - AvalRS‖ (2008–2011). MEASURING WEAK LAYER FORMATION Snow variables of importance for weak layer formation and which might be retrieved by using remote sensing data include snow grain size, snow surface temperature and snow wetness. The temporal development of these variables will, under special conditions, also be essential, like, for example,the formation of surface hoar. Such a surface phenomenon could later result in a weak layer within the snow pack. The project carried out for NSC demonstrated that snow surface properties, related to the subsequent formation of weak snow layers in __________________________ * Corresponding author address:

Rune Solberg Norwegian Computing Center P.O. Box Blindern N-0314 Oslo, NORWAY tel: +472285 2500 fax: +472269 7660 email: rune.solberg@nr.no

the snowpack, could be measured in satellite data of moderate resolution during the accumulation season (Solberg et al., 2009; Solberg et al., 2010). We applied the MODIS sensors (on board the Terra and Aqua satellites) for snow parameter retrieval. In situ measured surface snow grain characteristics were compared to snow grain characteristics as derived from multispectral data from the MODIS satellite sensor. The study showed that parallel in situ snow measurements and snow analyses exploiting data from MODIS are possible for the selected test sites in Norway. We found two cases where the satellite-retrieved snow surface temperature and snow grain size evolution indicated formation of surface hoar. Unfortunately, we lack in situ measurements for those cases, while a skiing tourist observation at the same time in a nearby mountain area confirmed extensive surface hoar formation. As the weather conditions where the same in both regions, this is a strong indication that the remote sensing results were correct. DETECTING AVALACHES The avalRS project carried out a pilot study and demonstration showing that pattern recognition methodology could be used for the detection of avalanches using very-high resolution (VHR) optical satellite data (Frauenfelder et al., 2012 Larsen et al., 2011). We explored the use of imagery from highresolution (HR) and very high-resolution (VHR) satellite sensors. HR included SPOT 2 and 4, and VHR included Quickbird and test data from an airborne optoelectronic pushbroom scanner (data courtesy of Leica-Geosystems, Heerbrugg, Switzerland and Y. Bühler, SLF, Switzerland). Only VHR imagery (≤ 2.5 m spatial resolution) gave suitable results. Similar VHR sensors are currently available on, e.g., the WorldView, Ikonos, Orbview, GeoEye, SPOT-5 and Pléiades satellites. The key part of the detection algorithm is a texture segmentation step, which distinguishes the avalanches from other objects such as smooth and rugged snow, trees and rock. Two


R. SOLBERG, R. FRAUENFELDER S.Ø. LARSEN, A.-B. SALBERG NORWEGIAN COMPUTING CENTER (NR) NORWEGIAN GEOTECHNICAL INSTITUTE (NGI)

ADVANCES IN AVALANCHE FORECASTING, PODBANSKÉ

Per year worldwide have been reported as a direct result of snow avalanches (Schweizer, 2008). We have presented results of experiments using satellite remote sensing to measure the formation of surface hoar and detection of recent avalanches.Strong evidence was found for observations of surface hoar based on the combination of snow surface temperature and snow grain size development. Further work aims at establishing a snow grain evolution model. The model will be used as an input to the avalanche forecasting model.

FIGURE 1.Two avalanches in Hellesylt, Norway, acquired by Quickbird on 16 April 2005. Left: avalanches outlined manually in blue on the panchromatic image. Right: detected segments are overlain in pink.

different approaches are investigated: a method based on grey‐ level co‐ occurrence matrices (GLCM), and a method based on directional filters. The algorithms are developed and trained on a Quickbird image of a Norwegian mountain area which contains several avalanches. The segmentation algorithms detect parts of all avalanches. The directional filter method was also tested and validated on another Quickbird image, covering a different scene in Norway. The GLCM approach gives a higher rate of false detections than the directional filters approach, but maps the outline of the avalanches better. A brief demonstration of feature extraction shows that context and shape of detection objects may provide important information to further enhance the performance by reducing the number of false detections and refining the outline. From this case study, we believe that avalanche mapping in VHR optical images is possible in general. CONCLUSIONS Avalanches rank among the most significant natural hazards in the snow-covered mountains of the world. Avalanches endanger not only transport links and settlements, but also tourists who visit ski resorts and back-country mountain regions during wintertime. During recent decades, an estimated 250 causalities

In our study of detecting avalanches, we have showed that automatic detection and mapping of occurred avalanches in VHR optical imagery is possible. The study also showed that such data could be used to add important information to validate avalanche forecasts, something which could lead to significantly improved forecasts. We think that an operational service for such inventory and verification purposes could be turned into practice once the remaining algorithm challenges have been overcome. REFERENCES Frauenfelder, R., Solberg, R., Larsen, S.Ø., Salberg, A.-B., Bjordal, H. 2012. Remote-Sensing Derived Avalanche Inventory Data. Proceedings of the International Snow Science Workshop (ISSW), 17-21 Sept. 2012, Anchorage, USA.Contribution no. 1019. Larsen, S.Ø., Salberg, A.-B., Solberg, R., 2011.Evaluation of automatic detection of avalanches in high resolution optical satellite data. Results from the ESA avalRS project‘s feasibility study on automated avalanche detection. NR Note, SAMBA/23/11, 5 July 2011. Schweizer, J., 2008. Snow avalanche formation and dynamics, Cold Reg. Sci. Technol., 51, 153–154. Solberg, R., Frauenfelder, R., Koren, H. and Kronholm, K., 2009. Could retrieval of snow layer formation by optical satellite remote sensing help avalanche forecasting? Presentation of first results.International Snow Science Workshop (ISSW 2010), 27 September-2 October 2009, Davos, Switzerland. Solberg, R., Koren H. &Wangensteen, B, 2010. Remote sensing of snow characteristics for avalanche warning – ―Snøskred‖ project results from 2008-2009. NR Note, SAMBA/09/10, 22 March 2010.


ROMAN SITKO FACULTY OF FORESTRY TECHNICAL UNIVERSITY IN ZVOLEN

ADVANCES IN AVALANCHE FORECASTING, PODBANSKÉ

Avalanche forecasting system for the forest managers common aspects R. Sitko* Faculty of forestry, Technical University in Zvolen, Slovakia COMMON BASES AND DIFFERENCES What is avalanche forecasting (AF) system for the forest managers and which are common bases and differences compared to regular AF systems? The main aim of the forest managers in the mountains endangered by avalanches is to keep forest ecosystem sustainable active to protect landscape, lives and properties. The very important aspect is that trees, which are in the center of the forester interest can not to be moved in case of high avalanche danger as the people can. So the most important question for forester is: Where can the forest inhibit the avalanche triggering (primary avalanche role of the forest) and where can it brake already triggered avalanches (secondary avalanche role of the forest)? Because of that, the main factors influencing the efficient forest managers decisions are the long-term so-called sustained factors: topographic, climatic and land cover. Less important are those factors, which vary in time, like meteorological ones. METHODS The model area is situated in the part of West Tatra Mountains called Tichá valley. The territory has a high-mountain character with a rugged terrain relief with altitudes ranging from 1,100–2,052 m above sea level. For the purpose of input identification, the topographic variables of the terrain were derived from digital elevation model (DEM). The slope, slope length, surface curvature, contributing area and altitude were chosen as the most important terrain variables for avalanche control (AC) role of the forest evaluation. Land cover was classified from Ikonos satellite data. Work flow of the avalanche control role of the forest evaluation display Figure 1. Quantification of AC role of the forest was based on the assessment of avalanche triggering plausibility (strength of evidence) and avalanche track area measures for their maximum run-out distance allocations. Expert __________________________ * Corresponding author address:

Roman Sitko Faculty of forestry, Technical university in Zvolen T.G. Masaryka 24 960 53, Zvolen, SLOVAKIA tel: +421 45 5206 294 e-mail: sitko@tuzvo.sk

system for release zone identification was developed. NetWeaver logic engine was utilized for knowledge base definition. The fuzzy membership functions were defined to express plausibility for the criteria (arguments) of avalanche triggering (topic). The maximum run-out distance was modeled by statistical Alpha-beta model (Barbolini et al., 2000). Input identification:

DEM analyses

actualization

IKONOS images, land cover classification

Quantification: NetWeaver

Map of observed avalanches

Alpha-beta model Maximum run-out distance modeling (S1, S2, S3)

Release zones identification

Evaluation:

Field survey

EMDS

Primary AC scenarios (S1, S2, S3)

Secondary AC scenarios (S1, S2, S3) Final AC scenarios (S1, S2, S3)

AC potential (FP=S3-S1)

Zoning:

AC effect (FE=S3-S2)

GIS Priority zone of management support

FIGURE 1. Scheme of work flow for avalanche control role of the forest evaluation and zoning (AC-avalanche control, S-scenario, FP-functional potential, FE-functional effect)


ROMAN SITKO FACULTY OF FORESTRY TECHNICAL UNIVERSITY IN ZVOLEN

ADVANCES IN AVALANCHE FORECASTING, PODBANSKĂ&#x2030;

FIGURE 2. Knowledge base for release zones identification

In general, the avalanche processes were quantified for the following three scenarios: optimal state of forest (S1) characterized by natural (optimal) tree species composition, real state of forest (S2) given by the actual tree species composition, none impact of the forest on the quantified process (S3). The next step is evaluation of AC role of the forest for all three scenarios. Primary AC role of the forest was evaluated utilizing the knowledge base for release zone quantification. Secondary AC role of the forest was evaluated by the plausibility that forest can brake the avalanche. The assessment of knowledge bases was executed by Ecosystem Management Decision Support (EMDS) software. Finally, the final AC role of the forest was assessed and AC potential was evaluated as a difference between criterion score of scenarios S3 and S1 as well as S3 and S2 for AC effect evaluation. Allocation of appropriate forest priority zones for management support is the most important aim of applied methodology. It follows up the previous step by finding out the differences between the AC potential and AC effect (FP â&#x20AC;&#x201C; FE) Figure 1. The tools of GIS software (ArcGIS) were utilized for this purpose. For more details see Sitko, Scheer (2012).

The fuzzy membership function for six sustained factors of avalanche triggering has been designated: altitude, slope, land cover, canopy density, profile and planar curvature. The membership function definition was based on knowledge and data of nature snow pouring and snow layer accumulation by cooperation with Avalanche Prevention Center in Jasna (Slovakia). The factors (arguments) have been connected by AND operator, which inside of EMDS software brings so-called by minimum value biased weighted average (Reynolds et al., 2002). Testing of the knowledge base has shown this algorithm as sufficient for release zone identification. The map result of knowledge base assessment presents Figure 3. The plausibility values within the interval (-1, 1) reflect the strength of evidence for avalanche triggering. Red areas with -1 represent no support of avalanche triggering, 0 the undetermined degree of support and green areas with 1 represents the maximum degree of support for avalanche triggering. The values between -1 and 0 express decreasing strength of evidence for no support of avalanche triggering, values between 0 and 1 increasing one for full support.

RESULTS AND DISCUSION This section is focused on results of quantif ication phas e of intr oduc ed methodology. The scenario S2, which was used for calibration of release zone identification knowledge base (Figure 2) and maximum run distance modeling, is in thecenter of interest.

Based on the results of the knowledge base, the thematic map of release zones was created. The threshold values of plausibility and minimum area of release zone have been determined. Allocated release zones display Figure 4. Dark blue areas were classified as a release zones by the highest Kappa index of agreement with mapped release zones.


ROMAN SITKO FACULTY OF FORESTRY TECHNICAL UNIVERSITY IN ZVOLEN

FIGURE 3. The map of plausibility for avalanche triggering (Uzemie.shp – model area, lav_drahy – mapped avalanche tracks, L_actual – plausibility for real avalanche triggering)

FIGURE 4 Modeled vs. mapped avalanche tracks (Uzemie.shp – model area, lav_drahy – mapped avalanche tracks, Odtrh_zona – release zone, LD_model – modeled avalanche tracks)

There was modeled maximum run-out distance for identified release zones. The statistical Alpha-beta model derived from the data of 17 observed avalanche tracks has acquired the coefficient of determination 0.98 and its standard error was ±1.12°. Those results the most correspond to Austrian model presented by Barbolini et al. (2000). The final result of avalanche tracks modeling displays Figure 4. Blue areas represent the modeled avalanche tracks. The Kappa index of agreement for modeled versus mapped avalanches acquired 0.58, for class of avalanche tracks 0.82.

ADVANCES IN AVALANCHE FORECASTING, PODBANSKÉ

It suggests to overestimation of modeled avalanche area and the highest overestimation was spotted at the stands over timberline. Due to necessity of responsible data for the stands over timberline, there is not only explanation of that result. On the one hand it could be interpreted as an improvement of mapped avalanche tracks over timberline for which was not possible to detect their tracks properly from airborne or satellite images. On the other hand it could be the consequence of low representativeness of sample data for model derivation. As a sample data were utilized big avalanches intervening under timberline usually as far as mountains valley. In case of small avalanches occurred over timberline it could not be representative sample data for model derivation. In such a case it would be necessary to stratify the model area to avalanches under and over timberline and to derive one model for every stratum. Based on the experience of presented study we can conclude: EMDS and NetWeaver logic engine utilizing the fuzzy set membership functions is suitable tool for evaluation of avalanche triggering due to probability character of avalanche phenomena. It is necessary to acquire and use the responsible data of avalanches over timberline to confirm the wider utilization of statistical Alpha-beta model at the model area. In accordance with Barbolini et al., (2000) we suggest to utilize the advantages the both kind of avalanche models, statistical as well as physical based ones. ACKNOWLEDGEMENT This study is the result of the implementation of the project: Centre of Excellence ―Decision support in forest and country‖, ITMS: 26220120069, supported by the Research & Development Operational Programme funded by the ERDF. REFERENCES Barbolini, M., Gruber, U., Keylock, C.J., Naaim, M., Savi, F., 2000: Application of statistical and hydraulic-continuum dense-snow avalanche models to five real European sites. Cold Regions Science and Technology 31, pp. 133–149. Reynolds, K.M., Rodriguez, S., Bevans, K., 2002. EMDS 3.0 User Giude: U.S. Department of Agriculture, Forest Service Sitko, R., Scheer, Ľ., 2012: Decision support and avalanche control role of the forest evaluation. In: Implementation of DSS into the forestry practice. The first international scientific conference, Zvolen, Slovakia, May 10 – 12th, 2012, (in press)


P. CHRUSTEK, N. KOLECKA, Y. BÜHLER JAGIELLONIAN UNIVERSITY, KRAKÓW ANNA PASEK FOUNDATION, BĘDZIN WSL INSTITUTE FOR SNOW AND AVALANCHE RESEARCH SLF, DAVOS

ADVANCES IN AVALANCHE FORECASTING, PODBANSKÉ

Snow avalanches mapping – evaluation of a new approach Paweł Chrustek1,2,* ,Natalia Kolecka1 ,Yves Bühler3 Jagiellonian University, Institute of Geography and Spatial Management, Kraków, Poland 2 Anna Pasek Foundation, Będzin, Poland 3 WSL Institute for Snow and Avalanche Research SLF, Davos, Switzerland

1

ABSTRACT Recent snow avalanche hazard mapping tools and procedures offer methodsto improve the accuracy and reliability of risk and hazard localization. The validation of numerical mass movement models mainly depends on recorded historical avalanche data setssuch as avalanche outlines and release volumes. These data sets are often unavailable or of an unknown accuracy. Avalanche characteristics such as release area, flow height and flow path, runout distance and total amount of released snow mass are essential parameters for proper calibration and evaluation of numerical simulation tools. Incorrectly calibrated models can influence decisionsmaking which directly affects human safety. The acquisition of high quality data regardingobserved avalanche events is often hindered by the high risk permanently present in avalanche terrain. Thispaperdescribes a promising methodbased on photogrammetry and computer vision and also introduces AVALMAPPER software that allows using a single terrestrial photograph with unknown exterior and interior orientation parameters toaccurately map avalanche outlines. We evaluate this methodby comparingits results with GPS measurements made in the field. We discuss the optimization of measurement efficiency, costsand human safety. INTRODUCTION Each year snow avalanches cause a great number of accidents in mountainous areas. This force of nature brings not only casualties, but also significant forest and infrastructure damage. Rescue statistics show that the common cause of avalanche accidents is due to a difficulty in spatial risk factor evaluation. This problem can be related to the lack of knowledge and experience, as well as objective causes (e.g. weather conditions), __________________________ * Corresponding author address:

Pavel Chrustek Jagiellonian University Institute of Geography and Spatial Management ul. Gronostajowa 7 30-387 Krakow, POLAND tel.: + 48 32 761 53 80 fax.: + 48 32 761 53 77 e-mail: p.chrustek@annapasek.org

which may affect both amateurs and specialists. For this reason, it is very important to supplement snow avalanche education with the knowledge from the field of Geographic Information System (GIS) and remote sensing of environment that provide methods and tools for improved spatial risk and hazard location (Chrustek 2009). Avalanche hazard mapping tools and procedures developed by avalanche specialists from all over the world offer an increasing number of methods for providing more accurate risk and hazard localization. Numerical avalanche dynamics models like RAMMS (Christen et al. 2010), SAMOS (Sampl and Zwinger 2004) or ELBA+ (Volk and Kleemayr 1999, Sauermoser and Illmer 2002) coupled with GIS have become an essential part of snow engineering and hazard mapping studies (Christen et al. 2010). Unfortunately, a constant fundamental problem in the science of snow mass movements is to record and document occurring events. Postevent mapping is helpful to determine release areas, volumes of the released masses, runout distances and inundation areas. This kind of data is necessary for developing and evaluating new mitigation methods and tools.Avalanche mapping is also important for risk assessment verification and accident documentation. Beside conventional avalanche data (altitude, aspect, slope, size, etc.) a general outline marked on maps should be a part of each avalanche survey (Meister and Jeller 2009). On a global scale, the documentation of avalanches remains sparse and therefore incomplete, and itsaccuracy is unknown. Currently, detection and mapping of observed avalanches relies mainly on isolated observations acquired by individual experts under field conditions. Quite often, only avalanches causing accidents or resulting in heavy damages are mapped (Bühler et al. 2009). At certain locations, such as Davos in Switzerland, long-term records of well documented events exist. However, because of the changing climate and missing records, historical data may not show the complete picture of the current threats.


P. CHRUSTEK, N. KOLECKA, Y. BÜHLER JAGIELLONIAN UNIVERSITY, KRAKÓW ANNA PASEK FOUNDATION, BĘDZIN WSL INSTITUTE FOR SNOW AND AVALANCHE RESEARCH SLF, DAVOS

A common lack in high quality data of localized snow avalanche releases and depositions is more often caused by the high risk permanently present within avalanche areas (especially in the release zone), rather than limited availability of measurement devices. This is the reason that the most popular method used by experts is manual mapping based on remote observations and topographic maps. However, according to opinions of avalanche experts this method requires high skill levels, and very often leads to numerous discrepancies between field measurements and drawn extents (Meister and Jeller 2009). Traditional hand-held Global Navigation Satellite System (GNSS) measurements allow cheap, easy and accurate mapping of avalanches, but they are time consuming and often of restricted applicability due to avalanche danger. The quality of these measurements strictly depends on the available satellite signal which may be significantly reduced in the complex terrain (Chrustek et al. 2010). Remote and high resolution survey isenabled by integrating the two technologies:Light Detection And Ranging(LiDAR)and GNSS (e.g.Deems and Painter 2006;Jörg et al. 2006;Vallet 2008;Prokop et al. 2008). Unfortunately, the most important disadvantage of this method is high price and limited repeatability. For many years, obtaining this kind of data has been impossible for many operational and researchbudgets. Application of these technologies also requires large amounts of time for detailed measurement planning, and qualified staff that may operate system and process data (Deems and Painter 2006; Prokop 2009). Single terrestrial photographs can also be used as a source of valuable geographical information. Recently, Aschenwaldet al. (2001) and Corripio (2004) presented an approach to incorporate a single terrestrial photograph into geographical analysis. Their methods, however, employed photographs taken from known or measured locations, and this subsequently limits potential data sources to a new or well-documented set of photographs. An interesting approach in the avalanche outline mapping context was presented by Meister et al. (2009). They used a digital ―Atlas of Switzerland 2‖ (2004) and digital terrestrial pictures containing Global Positioning System (GPS) coordination and azimuth parameters and based on the digital elevation model(DEM) panorama and tools for adjusting digital pictures, avalanche outlines were drawn on screen.

ADVANCES IN AVALANCHE FORECASTING, PODBANSKÉ

Our aims were to develop an innovative method and create the AVALMAPPER software which allow to map an avalanche extent using a single terrestrial photograph with unknown exterior and interior orientation parameters. Our idea is based on creatinganorthoimage from the terrestrial photograph by means of photogrammetric and computer vision rules, and subsequent visual interpretation. METHODS Theoretical background The proposedmethodology contains three principal steps. First, the photograph needs to be oriented in a global coordinate system to obtain camera position and to perform function mapping of 3D points into a 2-dimensional (2D) image. Next, the viewshed analysis is performed to determine parts of terrain that are not visible from the particular camera position. In the third step the visible DEM points are projected into the photograph by means of the mapping function, to obtain color information out of the photo. All mentioned steps were coded using Python programming language and linked with the Graphical User Interface (GUI) developed for this task (Fig. 1). To orientate a terrestrial photograph, the Direct Linear Transformation (DLT) method is applied (Abdel-Aziz and Karara 1971;Luhmann et al. 2006; Kraus 2007). The DLT establishes the relationship between the 2D image coordinates and the 3D object coordinates, using projective transformation rules and ground control points (GCPs).The control points must have image and global coordinates measured to compute the camera orientation parameters. A minimum of 6 points are necessary to solve the DLT;however, to cope with images from noncalibrated camera or scanned old photographs, more GCPs are necessary to obtain accurate results. Global points‘ coordinates can be surveyed in the field, most likely with GNSS device, or determined by examining existing aerial orthophotomaps and elevation models. Eleven DLT parameters can be computed from linear equations, so no approximate values of the unknowns are required (Luhmann et al. 2006;Kraus 2007).

x

L1 X L9 X

L2Y L3 Z L L10Y L11 Z 1

x

L5 X L9 X

L6Y L7 Z L L10Y L11 Z 1

(1)


P. CHRUSTEK, N. KOLECKA, Y. BÜHLER JAGIELLONIAN UNIVERSITY, KRAKÓW ANNA PASEK FOUNDATION, BĘDZIN WSL INSTITUTE FOR SNOW AND AVALANCHE RESEARCH SLF, DAVOS

whereL-1...L11are DLT parameters, x, y are image coordinates, andX, Y, Zare object coordinates in the ground coordinate system. From L1...L11, the coordinates of the projection center: X0, Y0, Z0 (camera location) are computed, as they are necessary to determine visible and hidden parts of terrain (Luhmann et al. 2006). Knowing the coordinates of the projection center, the invisible parts of terrain should be eliminated in order to avoid projecting hidden points into the photograph. This goal is achieved by applying the viewshed analysis to the raster DEM. The approach introduced by Wang et al. (2000) is utilized, which uses the concept of reference planes. A traditional orthophoto is created when relief displacements are removed from the original image (Kraus 1992; Novak 1992; Mikhail et al. 2001;Okeke 2001;Kraus 2007). It is done using DEM, whereby the interior and exterior parameters of the image are used. There are two ways to conduct such rectifications: forward and backward projection, the latter was used in this work (Fig.2). Each pixel is projected back to the image space using the DLT equations (1), and then attributed with the image luminance value, obtained by resampling. Output from this procedure is the orthophotograph, saved as a TIFF file and georeferencingparameters stored in a TFW world file. The essential step for DLT quality assessment is the comparison of GCPs coordinates measured in the photograph with GCPs global coordinates projected onto the photo. The residuals (optical error) are computed as follows:

ADVANCES IN AVALANCHE FORECASTING, PODBANSKÉ

FIGURE 1. Graphical User Interface of AVALMAPPER software

(2)

wherevx, vyare differences between measured and computedimage coordinates of thecontrol points by x and y, and vxyis atotal optical error.

FIGURE 2. .Backward projection during orthophoto creation in the aerial case

The output orthophotograph is visually investigated. It involvesboth checking the general appearance of the image and comparing it with existing data, e.g. traditional (aerial) orthophotomap. Usually it can be seen if such an image contributes some new, additional information, or if it rather contains smudges or other artifacts. Finally, computed orthoimages become a base layer for manual or automatic vectorization process. Correct


P. CHRUSTEK, N. KOLECKA, Y. BÜHLER JAGIELLONIAN UNIVERSITY, KRAKÓW ANNA PASEK FOUNDATION, BĘDZIN WSL INSTITUTE FOR SNOW AND AVALANCHE RESEARCH SLF, DAVOS

interpretation of the processed image strictly depends on the operator experience, the selected map scale and assumed automatic classification method. Experimental work The study areaswere located in the Polish Tatra Mountains and the eastern part of the Swiss Alps around Davos. To test the approach two different locations were chosen (Fig. 3): DorfbergMountain (DB) – a 2592 x 3872 pixels digital photograph, taken with noncalibrated digital SLR Sony A100 camera equipped with 10.8 millions pixels CCDmatrix and a zoom lens 17-50 mm. RGB colors show south-easternexposed slopes of the Dorfberg Mountain and surroundings (picture from 9thof February 2010). The photograph recorded a few small and medium size snow avalanches,the biggest one in the central part of the picture (aproximately 220 m long) was also measured by GPS Trimble GeoXH device (with decimeter accuracy). Metadata are not available from the EXIF file. GoryczkowaCzuba (GC) – a 1013 x 661 pixel copy of the scanned analog photograph, R G B c o l o r s s h o w CzubaGoryczkowaMountain and avalanche rescue action that took place on 11th January 1985. Part of medium size avalanche (approximately 600 m long) visible on the picture was released by a tourist. It killed one person and caused injuries to another. DEM for the GC test sites was available as the TIN (Triangulated Irregular Network) modelbased on contour lines with 5 m intervalsdigitized from topographic maps 1:10000), mass points and hardlines. It was converted to the raster format with spatial resolution of1 m. DEM for the DB site, with 2 m resolution, was produced from aerial images. GCPs were measured in ArcGIS software, on a basis of orthophotomaps and DEM. 9 GCPs for

ADVANCES IN AVALANCHE FORECASTING, PODBANSKÉ

each test site were found resulting in various optical errors (Table 1). TABLE 1. Comparison of GCPs coordinates measured in the photograph and GCPs global coordinates projected into the photo Test Sites coordinates mean residual [pixels] standard deviation [pixels] optical error by x and y [pixels] total optical error [pixels]

DB

GC

x

y

x

y

-2.57

0.44

-1.53

0.80

4.18

3.30

2.99

7.04

4.70

3.30

3.29

6.91

5.66

10.98

Created orthoimages became base layers for vectorization. Avalanches outlines were manually vectorized on screen using GIS software. RESULTS Despite the low contrast of the snow surface, both release and deposit zones are visible in the processed orthoimage from the DB test site (Fig. 4 d). A few tests showed that more detailed shape of the avalanche outline may be obtained when the outline is marked on the input photograph. The vectorization process on the orthoimage then becomes much easier. Figure 4 c shows the comparison between vectorized and field measured data. This example strongly demonstrates that high accuracy avalanche outline mapping based on traditional terrestrial photograph is possible. The measured vertical differences between the outlines do not exceed 15 meters. It is worth mentioning that vertical distance between camera and the avalanche was over 1500 m. Based on this fact, we can say that the most important advantage of this method is that it works very well without accessing dangerous areas. It also saves time, cost and effort. Very promising results were obtained when analyzingorthoimage processed from the old scanned photograph in the CG test site (Fig. 5). This example showed that processing of an

FIGURE 3. The test images with marked GCPs: DB (left),GC (right)


P. CHRUSTEK, N. KOLECKA, Y. BÜHLER JAGIELLONIAN UNIVERSITY, KRAKÓW ANNA PASEK FOUNDATION, BĘDZIN WSL INSTITUTE FOR SNOW AND AVALANCHE RESEARCH SLF, DAVOS

ADVANCES IN AVALANCHE FORECASTING, PODBANSKÉ

FIGURE 4. DB site orthoimage on the traditional orthophotomap background (b) and vectorized avalanche outline (c) compared with field GPS measurements (dashed black polylines)

also gives a chance to compute errors in the camera orientation and assess its influence on further processing of the photograph. Similarly to Aschenwald et al. (2001) and Corripio (2004) in the proposed method GCPs need to be collected. This is the most time-consuming and problematic part of the georectification process. What affects accuracy of the GCPs measurement are photograph and orthophotomap resolution and DEMs accuracy or – in case of field survey – GNSS device / measurements accuracy. It is difficult to find reliable GCPs on old photographs, whether in the orthophotomap or in the field due to old photograph with lim ited identification of reliable locations of the GCPs is possible (most of characteristic terrain features were completely covered by the snow). Because of the lower resolution of the source photograph, the vec torization of the vis ible avalanche outline was also much easier when the outline was marked on the source image (Fig 5 b). The avalanche boundaries visible on the orthoimage processed from the raw picture are not so clear (Fig. 5 c), and the whole outline of the avalanche was not visible. As the documentation prepared after the rescue action contains a detailed description, it can help to identify precise location of the runout limits and draw them on the map in the next step. DISCUSSION AND CONCLUSIONS Our paper presents the method that minimizes necessary input information and replaces approximate data with mathematical calculations, based on the DLT method that has several advantages as compared to earlier works by Aschenwald et al. (2001) and Corripio (2004) (Table 2).First, mathematical computations seem to be more reliable than error-prone visual estimation; another DLT advantage is that it works with terrestrial, oblique and aerial photographs, making this method multi-purpose. Mathematical formulae

FIGURE 5. CG site orthoimages on the traditional orthophotomap background (c) and 3D visualization of the draped ortoimage on the DEM (a). D: orthoimages with visible avalanche outline (continous line) drawn earlier on the source image

changes in landscape features and snow coverage when analyzing winter images. The problem applies also to very steep slopes that look different fromthe terrestrial and aerial perspectives or can be hardly accessible for surveys. Nevertheless, as the DLT method needs only a minimum of six GCPs, so it is highly probable that they can be found, as proven in the GC test site.


P. CHRUSTEK, N. KOLECKA, Y. BÜHLER JAGIELLONIAN UNIVERSITY, KRAKÓW ANNA PASEK FOUNDATION, BĘDZIN WSL INSTITUTE FOR SNOW AND AVALANCHE RESEARCH SLF, DAVOS

ADVANCES IN AVALANCHE FORECASTING, PODBANSKÉ

TABLE 2. Benefits in comparison with Aschenwald et al. (2001) and Corripio (2004) Methods

Aschenwald et al. 2001

Corripio 2004

method proposed in this paper

Theory

photogrammetry (space resection)

computer vision

photogrammetry (Direct Linear Transformation)

Input data

camera and target coordinates DEM – list of XYZ-points in text file visibility map – (0,1) text file – in the same order as DEM points GCPs – min. 15 photograph camera focal length

camera and target coordinates DEM – list of XYZ-points in text file visibility map – (0,1) text file – in the same order as DEM point GCPs – min. not defined photograph camera focal length

DEM – list of XYZ-points in text file visibility map – (0,1) text file – in the same order as DEM points GCPs – min. 6 photograph

Obtaining camera and target coordinates

fixed camera position image center from diagonals of the photo camera and target coordinates – from orthophotomap, by approximation

fixed camera position target by approximation

omitted

GCPs utilization

for quality assessment for camera position optimization – iterative checking of different positions in 1m grid around the approximate coordinates

Additional corrections

image size reduced to a smaller extent

for quality control in the photo distance between each GCP measured in the photo and backprojected GCP is checked trial and error camera and target optimization roll of the camera around the viewing direction axis Earth curvature and light refraction – included but can be neglected due to marginal influence on the results

In comparison to Aschenwald et al. (2001) and Corripio (2004), an additional improvement of the method isthe implementation of the viewshed analysis into the final procedure that has not been proposed until now. The algorithm of Wang et al. (2000) was successfully programmed and tested to assess its quality, with the results almost identical to viewshed procedures widely used in proprietary GIS software. All the benefits of our method in comparison with method proposed by mentioned authors were presented in Table 2.

FIGURE 6. Results of the avalanche mapping in DB test site using AVALMAPPER software and different input DEMs: continuous black line – 2 m resolution DEM produced from aerial high resolution digital images, dashed black line – 25 m Swisstopo DEM, dashed dot white line – 27 m ASTER GDEM, white continuous polylines show GPS measurements

for photo orientation with DLT for quality assessment

any photo / image can be processed rolls and tilts – included in DLT

The positional accuracy and quality of the final result largely depends on the input DEM quality (Fig. 6). As elevation models of highmountainous areas produced from aerial photographs have major inaccuracies in a very steep or shaded terrain, these errors will be propagated into the derived data (Foote and Huebner 1995; Krupnik 2003). Analyzed examples show that the processing of terrestrial images from different sources and different times, including old analog photographs, is possible. With a good performance of the algorithm, high-resolution orthophotographs can be easily obtained, allowing great capabilities for the visual interpretation in a standard GIS software. Also for shaded slopes, results mayprovide as good radiometrical information as RGB aerial photographs. The unquestionable advantage of such data istheir low price. They can also be obtained almost anytime, independently of the season or equipment. Bad weather conditions can be an obstacle, but it is a common feature of many data-collecting techniques, e.g. laser scanning. Imaging at night is impossible as well. A drawback of terrestrial orthophotos isthat some parts of the terrain may not be visible; however, itcould be easily compensated with multiple terrestrial photographs. The potential field of applications is wider than avalanche mapping and may be used to gather information on other natural hazards in difficult terrain such as debris flows, landslides and rock falls.


P. CHRUSTEK, N. KOLECKA, Y. BÜHLER JAGIELLONIAN UNIVERSITY, KRAKÓW ANNA PASEK FOUNDATION, BĘDZIN WSL INSTITUTE FOR SNOW AND AVALANCHE RESEARCH SLF, DAVOS

ACKNOWLEDGEMENTS We would like to express our gratitude to the Foundation for Polish Science for financial support of Paweł Chrustek. Performing the analyses was possible thanks to the VENTURES program organized by the Foundation of Polish Science and co-funded by the European Regional Development Fund under the Operational Program Innovative Economy 2007-2013.Natalia Kolecka is a grant holder of "Doctus"Programme.We also would like to express gratitude to our colleagues Marek Świerk from the Anna Pasek Foundation in Poland, for assistance in collecting field data, Wojciech Bartkowski and Jakub Radliński from the Volunteer Mountain Rescue Service (Górskie Ochotnicze Pogotowie Ratunkowe, GOPR), for assistance in collecting data and materials concerning historical avalanches in the Tatra Mountains. REFERENCES Abdel-Aziz YI, Karara HM (1971) Direct linear transformation from comparator coordinates into object space coordinates in close-range photogrammetry. In: Proc. of the ASP/UI Symposium on Close-Range Photogramm. American Society of Photogrammetry, Falls Church, VA, 1–18. Aschenwald J, Leichter K, Tasser E, Tappeiner U (2001) Spatio-temporal landscape analysis in mountainous terrain by means of small format photography: a methodological approach. IEEE Trans on Geosci and Remote Sens, Vol. 39 (2001), No. 4., 885-893. Atlas of Switzerland 2 (2004) [DVD or 2 CD-ROMs]. Swiss Federal Office of Topography, Wabern. Bühler Y, Hüni A, Christen M, Meister R, Kellenberger T (2009) Automated detection and mapping of avalanche deposits using airborne optical remote sensing data. Cold Regions Science and Technology (57) 2-3, 99 – 106. Christen M, Kowalski J, Bartelt P (2010) RAMMS: Numerical simulation of dense snow avalanches in three-dimensional terrain. Cold Regions Science and Technology (63) 1-2, pp. 1 – 14. Chrustek P, Biskupič M, Kolecka N (2010) Comparison of different methods for obtaining snow avalanche data. In: Ostapowicz K, Kozak J (eds) Conference Proceedings of the 1st Forum Carpaticum, Integrating Nature and Society Towards Sustainability, Cracow, Institute of Geography and Spatial Management, Jagiellonian University, Poland, 111. Corripio JG (2004) Snow surface albedo estimation using terrestrial photography. Int. J. of Remote Sens., 25 (24), 5705-5729.

ADVANCES IN AVALANCHE FORECASTING, PODBANSKÉ

Chrustek P (2009) Promotion of secure mountain exploration by the Anna Pasek Foundation.In:Schweizer, J.; Van Herwijnen, A. (eds) International Snow Science Workshop. 27 September to 2. October 2009, Davos, Switzerland. Proceedings. Birmensdorf, Swiss Federal Institute for Forest, Snow and Landscape Research, 495-499. Deems J, Painter T (2006) Lidarmeasurement of snow depth: Accuracy and error sources, Proceedings of the International Snow Science Workshop ISSW, Telluride, CO, USA, 1–6, 2006, 384–391. Foote KE, Huebner DJ (1995) Error, Accuracy and Precission. http://www.colorado.edu/geography. Accessed 28 October 2010. Jörg P, Fromm R, Sailer R, Schaffhauser A (2006) Measuring snow depth with Terrestrial Laser ranging system. Measuring snow depth with a terrestrial laser ranging system. In: Proceedingsof the International Snow Science Workshop, 1–6 October 2006, Telluride, Colorado. Telluride, CO, International Snow Science Workshop, 452–460. Kraus K (1992) Photogrammetry Fundamentals and Processes. Dummler Verlag, Bonn. Kraus K (2007) Photogrammetry: Geometry from Images and Laser Scans. Walter de Gruyter, Berlin. Krupnik A (2003) Accuracy Prediction for OrthoImage Generation. Photogramm Rec, 18(101), 41 –58. Luhmann T, Robson S, Kyle S, Harley I (2006) CloseRange Photogrammetry. Principles, Methods and Applications. Whittles Publishing, Scotland. Meister R, Jeller P (2009) Avalanche outline mapping with a digital GPS camera. In:Schweizer, J.; Van Herwijnen, A. (eds) International Snow Science Workshop. 27 September to 2. October 2009, Davos, Switzerland. Proceedings. Birmensdorf, Swiss Federal Institute for Forest, Snow and Landscape Research,101. Mikhail EM, Bethel JS, McGlone JC (2001) Introduction to Modern Photogrammetry. Wiley, New York. Novak K (1992) Rectification of digital imagery. PhotogrammEng and Remote Sens, 58(3): 339344. Okeke FI (2001) Review of Digital Image Orthorectification Techniques. http:// www.gisdevelopment.net. Accessed 28 October 2010. Prokop A, Schirmer M, Rub M, Lehning M, Stocker M (2008) A comparison of measurement methods: terrestrial laser scanning, tachymetry and snowe probing for the determination of the spatial snowdepth distribution on slopes. Annals of Glaciology, 49, 210-216.


P. CHRUSTEK, N. KOLECKA, Y. BÜHLER JAGIELLONIAN UNIVERSITY, KRAKÓW ANNA PASEK FOUNDATION, BĘDZIN WSL INSTITUTE FOR SNOW AND AVALANCHE RESEARCH SLF, DAVOS

Prokop A (2009) Terrestrial laser scanning for snow depth observations: An update on technical developments and applications. In: Schweizer, J.; Van Herwijnen, A. (eds) International Snow Science Workshop. 27 September to 2. October 2009, Davos, Switzerland. Proceedings. Birmensdorf, Swiss Federal Institute for Forest, Snow and Landscape Research., 192-196. Sampl P, Zwinger T (2004) Avalanche simulation with SAMOS. Annals of Glaciology (38), 393398. Sauermoser S, Illmer D (2002) The use of different avalanche calculation models practical experiences. In: International congress INTERPRAEVENT. 2:741–750 (in the Pacific Rim-Matsumoto, Japan). Vallet J (2008) High Precision LiDAR Mapping for Complex Mountain Topography, In: Proceedings of the Mountain Mapping and Visualization 6 th ICA Mountain Cartography Workshop, LenkimSimmental, Switzerland 249-254. Volk G, Kleemayr K (1999) ELBA - Ein GISgekoppeltes Lawinensimulationsmodell. An wendungen und P erspektiven, in Österreichische Zeitschrift für Vermessung und Geoinformation Heft 2 + 3, 84-92. Wang J, Robinson GJ, White K (2000) Generating vie wsheds without using sightlines. PhotogrammEng and Remote Sens, 66, 87-90.

ADVANCES IN AVALANCHE FORECASTING, PODBANSKÉ


I. MUDROŅ, J. RICHNAVSKÝ, P. CHRUSTEK M. BISKUPIČ INSTITUTE OF GEOINFORMATICS, VSB-TU OSTRAVA AVALANCHE W ARNING SERVICE, HZS - JASNÁ INSTITUTE OF GEOGRAPHY AND SPATIAL MANAGEMENT, KRAKOW

ADVANCES IN AVALANCHE FORECASTING, PODBANSKÉ

Snow Avalanche Risk Assessment in Territorial Planning Case of Settlement Magurka I. Mudron1,*, J. Richnavský2, P. Chrustek3, M. Biskupič2 Institute of Geoinformatics, VSB-TU Ostrava, Czech Republic 2 Avalanche Warning Service, HZS - Jasná, Slovakia 3 Institute of Geography and Spatial Management, Jagiellonian University Krakow, Poland 1

ABSTRACT Snow avalanche risk assessment is essential in urban and territorial planning of mountain regions. Almost every territorial plan in Slovak mountain area is missing a deeper or at least any snow avalanche risk study. In this paper is showed the importance of snow avalanche risk assessment, which is highlighted with a real case study of settlement Magurka. Snowfall records, snow avalanche cadastre, historical avalanche reconstruction using GIS, avalanche simulation, observation in terrain and scenarios with various initial conditions have been used to delineate the impacted area. The research evaluates the prevailing land use management regarding to avalanche hazard. The intensity of avalanche impact pressure is one of the important results. Special emphasis was taken on urban areas and infrastructure. 6 buildings are directly affected with snow avalanche risk. Additional 10 more buildings are potentially affected in case of even more catastrophic scenario (lower risk). In three cases is a part of road network affected. Territorial plan of Partizánska Ľupča, where settlement Magurka is southern part of it, comprehend clear deficiency leading out of the missing avalanche risk assessment. This study shows the importance and practice of risk assessment to avoid building new endangered structures in avalanche hazardous regions. INTRODUCTION Avalanches are natural processes in mountain regions. The most favourable sites are slope angles between 25°–55° (McClung and Shaerer, 2006). Settlements in mountain areas and underneath these slopes have always been confronted with natural hazards. The historical review revealed the fact, that Slovak settlements were no exceptions. The most catastrophic event happened in 1924, when a __________________________ * Corresponding author address:

Ivan Mudroň Institute of Geoinformatics, VŠB-TU Ostrava 17.listopadu 15/2172 708 33 Ostrava-Poruba, CZECH REPUBLIC tel: +420 774401029 e-mail: mud023@vsb.cz

huge avalanche released from the slopes of Kríţna and run through the settlement Rybô. It killed 18 people and the upper part of settlement was never restored. Snow avalanche damages did not stop over the last years. The chalet Rysy was fully destroyed in 2000 and a new one was built in the same avalanche prone area. There is lack of studies, even in mountain avalanche prone areas, which would regulate urban sprawl according to avalanche risk in Slovakia. Many settlements expanded towards the mountainsides, where no buildings had stood before. Plus more people are coming to the mountains and spending more time in the countryside. Thus the risk of catastrophic occurrence is significantly increased. The importance of dealing with snow avalanche risk is more topical than ever in Slovakia. The risk should be considered before taking any human activities in mountain areas. Best it should be legislatively incorporated in territorial planning. AVALANCHE HAZARD AND RISK ASSESSMENT In Slovakia there is no methodology to properly access the avalanche hazard and risk about the settlements. The studies are referring to criteria used in other countries or regarding to the aim of the study. Except for Iceland, avalanche hazard mapping in Europe is based on assessment without an explicit evaluation on the possible consequences on exposed elements, and thus, on risk (Cappabianca et al., 2008). In Austria and Switzerland the hazard zone‘s delineation is based on the estimated frequency of snow accumulation in release zones of avalanches. Subsequently a physical model is applied to calculate a corresponding run-out of avalanches. In Switzerland the limit of the hazard zones is located at the runout of an avalanche corresponding to snow accumulation with a frequency of 1/300 yr (Arnalds et al., 2004).For example in Norway the limit of the hazard zones is delineated by the 1/1000 year frequency of avalanches (Arnalds et al., 2004). The backbone for the risk basis of planning in Canada is the five part scale for avalanche size based on expected destructive potential. Standards for decisions in Canada are risk


I. MUDROŅ, J. RICHNAVSKÝ, P. CHRUSTEK M. BISKUPIČ INSTITUTE OF GEOINFORMATICS, VSB-TU OSTRAVA AVALANCHE W ARNING SERVICE, HZS - JASNÁ INSTITUTE OF GEOGRAPHY AND SPATIAL MANAGEMENT, KRAKOW

based, meaning average avalanche frequency (return period) and some measure of consequences are included (McClung, 2008). There are two common measures of destructive potential which can be incorporated into the consequence portion of risk. They are firstly predicted impact pressures and secondly destructive potential based on the five part Canadian system for sizing avalanches (McClung, 2008). The idea of using snow avalanche risk as a criterion for hazard zoning is discussed. The definition of acceptable risk should depend on location and intended human activities in this area. The loss of human lives should be a dominant factor when considering the acceptability of risk for the society (Grímsdóttir, 2008). Thus this paper focuses on built-up areas and infrastructures, or areas intended for development. Intended built-up areas are properties with any form of planning permission or properties found inside of a residential, commercial etc. zone although there have not been built any structures yet. Slovakia does not lack for space and proper building plots. Therefore it is important to reduce the risk to minimum. The risk based acceptability criteria represents any harmful snow avalanche hazard to the areas with built or proposed structures. So it includes also people living or remaining inside of them. Limit of acceptable risk or limit at which protective actions should be initiated is set very low to protect as much lives as possible and avoid almost all material damages. Table 1 is showing low avalanche hazard criteria for chosen European countries and our study. STUDY AREA The research is located in Magurka settlement, southern part of the cadastre of village Partizánska Ľupča (Figure 1 and 2). It is situated

ADVANCES IN AVALANCHE FORECASTING, PODBANSKÉ

TABLE 1. Low hazard mapping criteria, where T is the return period of snow avalanche and its impact pressure P

Country

Criteria

Swiss

T>300yrs (powder T>30yrs and P<3kPa)

France

T=100yrs and P<1kPa

Italy Our study

T>100yrs or (T=100yrs and P<3kPa) T>300yrs

just underneath the main mountain range of Nízke Tatry. The part of the area is situated on slopes with enhanced risk of snow avalanches between valleys of Ďurková and Močidlo. The urban sprawl in this touristic attractive location is noticeable over the last years, when new areas have been built up with cabins and recreational buildings (Figure 3).

FIGURE 1. Territorial (comprehensive) plan of Partizánska Ľupča. Settlement Magurka located in the south of the plan (source: Commune Partizánska Ľupča)

CASE STUDY Quantitative snow avalanche risk assessment expresses risk as the function of hazard and vulnerability and the particular element at this risk (IUGS, 1997). Both, hazard and vulnerability are functions of snow avalanche intensity. Avalanche hazard is mapped according to

FOUNDATIONS OF EXISTING STRUCTURES

FIGURE 2. Foundations of existing structures in Magurka


I. MUDROŅ, J. RICHNAVSKÝ, P. CHRUSTEK M. BISKUPIČ INSTITUTE OF GEOINFORMATICS, VSB-TU OSTRAVA AVALANCHE W ARNING SERVICE, HZS - JASNÁ INSTITUTE OF GEOGRAPHY AND SPATIAL MANAGEMENT, KRAKOW

ADVANCES IN AVALANCHE FORECASTING, PODBANSKÉ

used in this study is even worse unfavourable long-lasting snow and weather condition than in the case of the avalanche 1970 (see table 1). Modern numerical simulation software package RAMMS was used for avalanche reconstruction (Richnavský et al., 2011) and simulation of next scenarios. Correct identification of release zones is crucial for proper simulation. The release zone (e.g. Figure 6) were identified by the terrain parameters,

FIGURE 3. New built cabins in the research area (photo: Richnavský, J.)

the reconstruction of historical avalanche, prevailing attributes of locality, simulation of avalanche scenarios and snowfall records in the case of this study. The basis for snow avalanche hazard assessment was the historical avalanche Magurka 1970, when 4 people were killed. It can be categorised among the greatest avalanches ever observed in Slovakia (Richnavsky et al., 2011). Derivation of friction and other terrain parameters, delineation of potential release zones is based on reconstruction of this historical event (Figure 4 and 5). Other snow avalanche observations were taken into consideration too. Hazard mapping criteria

FIGURE 5. Maximal snow height according to avalanche reconstruction

FIGURE 4. Historical photos of avalanche release, track and deposit of avalanche in 1970 (photo: archive of APC)


I. MUDROŅ, J. RICHNAVSKÝ, P. CHRUSTEK M. BISKUPIČ INSTITUTE OF GEOINFORMATICS, VSB-TU OSTRAVA AVALANCHE W ARNING SERVICE, HZS - JASNÁ INSTITUTE OF GEOGRAPHY AND SPATIAL MANAGEMENT, KRAKOW

ADVANCES IN AVALANCHE FORECASTING, PODBANSKÉ

land cover and empirical data previous snow avalanche observations.

POTENTIAL RELEASE ZONE ZONE 50 - 70 M UNDER RANGE MOST CONCAVE SLOPES

Second component of risk refers to the vulnerability of people and buildings exposed to snow avalanches. To evaluate the potential damage of avalanche it is inevitable to know the maximal impact pressure. It can be 2 to 5 times larger than the average impact pressure of an avalanche (McLung, Schaerer, 2006) . Generally the largest impact pressure is in the front of the avalanche. Table 2 is showing the potential damage by avalanches. In this study we tried to avoid any harmful consequences to reduce risk as much as possible. Recreational and residential Land use is limited to point where the risk is reduced to minimum. So we excluded from land for development all hazardous areas, whatever they carried low or high risk. The highest risk of avalanche threat is in the valley of Viedenka (Figure 6). Together 16 structures are endangered and more than 100 m of paved road. In the other localities is also endangered power pole and another part of the paved road network. The following object should be protected against the avalanches or removed. The comprehensive plan of Partizánska Ľupča should take in consideration the existing snow avalanche risk before any tragedy should happen.

MOST CONVEX SLOPES FOREST COVER

FIGURE 6. Identification of potential release zone, example of Viedenka valley TABLE 2. Potential damage caused by average impact pressure of snow avalanche. (adapted by McLung, 2008)

Impact pressure 1 kPa 5 kPa 30 kPa 50 kPa 100 kPa 1000 kPa

Potential damage Break windows Push in doors Destroy wood-framed structures Damage steel power poles Uproot mature spurce Move reinforced-concrete structures

FIGURE 7. Excluded area from land for development classified by maximal impact pressure according to Viedenka valley release zone


I. MUDROŅ, J. RICHNAVSKÝ, P. CHRUSTEK M. BISKUPIČ INSTITUTE OF GEOINFORMATICS, VSB-TU OSTRAVA AVALANCHE W ARNING SERVICE, HZS - JASNÁ INSTITUTE OF GEOGRAPHY AND SPATIAL MANAGEMENT, KRAKOW

ADVANCES IN AVALANCHE FORECASTING, PODBANSKÉ

AVALANCHE DANGER ZONES YELLOW ZONE

BUILDING

BLUE ZONE

BUILDING IN DANGER

RED ZONE

CONCLUSION Most people dream of building their ideal home, so understandably, land with any form of planning permission is the most exclusive and thus also most expensive. Pristine natural environment of Magurka meets the dreams of some of these people. The area is located in favourable place for building recreation structures. Territorial (comprehensive) plan of Partizánska Ľupča, where settlement Magurka is southern part of it, opened new areas for urban sprawl. This document does not take in consideration snow avalanche risk, although the area partly lies in snow avalanche hazardous zone. New cottages were built in red zone, where avalanche impact pressure overreaches 30 kPa. The snow avalanche risk prevention should not be understood as tool for excluding favourable and attractive location but as a tool to avoid economical, environmental and life losses. ACKNOWLEDGMENT This article has been elaborated in the framework of the project SGS nr. SV51122M1/2101 and Anna Pasek Foundation. Impact of snow and terrain conditions in chosen Sudeto-Carpathian mountain ranges, on the size of avalanche hazard areas" under the VENTURES program organized by the Foundation of Polish Science and co-funded by the European Regional Development Fund under the Operational Program Innovative Economy 2007 - 2013. REFERENCES Arnalds, Jónasson, K., Sigurðsson (2004) Avalanche hazard zoning in Iceland based on individual risk in Annals of Glaciology, vol. 38, pp.285-290.

FIGURE 8. Avalanche danger zones according to maximal impact pressure reached by avalanche from Viedenka valley

Cappabinca, F., Barbolini, M. and Natale, F. (2008) Snow avalanche risk assessment and mapping: A new method based on a combination of statistical analysis, avalanche dynamics simulation and empirically-based vulnerability relations integrated in a GIS platform in Cold Regions Science and Technology, vol. 54, pp. 193-205. Grímsdóttir, H. (2008) Avalanche hazard and risk assessment in Iceland in Proceedings Whistler 2008 International Snow Science Workshop September 21-27, 2008, pp. 774-778. IUGS Working Group on Landslides, Committee on Risk Assessment (1997) Quantitative risk assessment for slopes and landslides — the state of the art in Landslide risk assessment: Proceedings of the International Workshop on Landslide Risk Assessment, Honolulu, Hawaii, USA, pp. 3-12. McClung, D. M. (2008) Risk-based Land-use planning in snow avalanche terrain in Proceedings of the 4th Canadian conference on geohazards: from causes to Management, Presse d l‘Université Laval, Québec, pp. 594. McCLung, D and Schaerer, P. (2006) Avalanche Handbook. The mountaineers book, Seattle, USA, pp.342. Richnavský, J., Biskupič, M., Mudron, I., Devečka, B., Unucka, J., Chrustek, P., Lizuch, M., Kyzek, F. and Matejíček, L., (2011). Using modern GIS tools to reconstruct the avalanche: A case study of Magurka 1970 in Sympozia GIS Ostrava 2011 Proceedings. Ostrava, Czech Republic, pp.175185. Commune Partizánska Ľupča. Online source, link: http://www.partizanskalupca.com/index.php? option=com_content&view=article&catid=56&id= 212&Itemid=58, September 2012.


ADVANCES IN AVALANCHE FORECASTING, PODBANSKÉ


R. JURAS, V. SPOUSTA, M. KOCIÁNOVÁ I. ŠPATENKOVÁ, J. PAVLÁSEK CZECH UNIVERSITY OF LIFE SCIENCES IN PRAGUE ŠPINDLERŮV MLÝN THE KRKONOŠE MOUNTAINS NATIONAL PARK, VRCHLABÍ

ADVANCES IN AVALANCHE FORECASTING, PODBANSKÉ

Water saturated avalanches in the Krkonoše Mts. Roman Juras1,*, Valerian Spusta2, Milena Kociánová3, Irena Špatenkova3, Jiří Pavlásek1 1 Czech University of Life Sciences in Prague, Prague, Czech Republic 2 Špindlerův Mlýn, Czech Republic 3 The Krkonoše Mountains National Park, Vrchlabí, Czech Republic ABSTRACT A database of avalanche occurrence in the Krkonoše Mts. and their basic description has been created since the winter season 1961/62. 869 records of different types of avalanches, classified after De Quervain et al. (1973), were gathered until the season 2005/2006 (Spusta et al. 2006). This database includes 243 cases (28 %) of mixed type - C7 (dry-wet) snow valanches and 82 cases (9,4 %) of wet snow avalanches - C2. Full-depth avalanches represent portion of 8 % and 32 % respectively in these categories. We have focused on four avalanches, which were classified as full-depth very wet valanches. The photographic documentation (e.g. FIGURE 3) proves that these valanches were triggered by presence of flowing water in the saturating zone and seem to be very similar to slushflows or slush avalanches described by Rapp (1960) from northern Sweden. Since the slushflows and slush avalanches have been known, a great attention has been paid to these phenomena, because the consequences of their occurrence are often destructive for the human property and human life (Hestnes 1985, Hestnes et andersen 1998, Perov 1998). Different types of water sources can saturate the snow and consequently trigger slushflows or slush avalanches, which can originate in drainage channels, brooks, hill mires, bogs or spring areas (Hestnes 1998). Intensive soaking of snow is also related to meteorological conditions like massive thawing period, rain or accumulation of water behind an obstacle like a block of ice, cumulated debris or depression. Pressure of water-saturated snow then exceeds critical value of tensile strength and basal friction which triggers the avalanche (Hestnes 1998, Onesti 1987). This moving mixture often contains some organic and inorganic material entrained from the slope. The flow reaches approximate density about __________________________ * Corresponding author address:

800 – 950 kg m-3, hence it is very powerful (Perov 1998). The authors distinguish slushflow and slush avalanche (also called ―water saturated avalanche‖). Slushflow contains much more water and is completely (full-depth) soaked by water. Slush avalanche is fully saturated only on the base, while the top parts of the profile could be either poor of liquid water or water free. We have come up with an analysis of possible reasons that might have caused the four mentioned slush avalanches in the Krkonoše Mts. from the localities of Malá Úpská Rokle, Pramenný Důl and Sněhová Strţ. Presence of springs in the trigger zones Small multiple slope springs saturated slush avalanches in Malá Úpská Rokle (e.g. FIGURE 1) on 2nd May 2002 and in Pramenný Důl (e.g. FIGURE 2) on 1st April 2006. At the locality of Malá Úpská Rokle, several permanent springs originate at an elevation of 1380 m a.s.l. on the steep eastfacing slope (angle ca. 30-40°) of Úpská Jáma cirque. The locality of Pramenný Důl is situated on the southern side of Luční Hora Mt. at ca 1400 m a.s.l. (similar angle), where a ca 50 m wide zone of springs (which is nearly the same width as the avalanche starting zone) can be found. Thawing water Water from melting snow can drain from developed slush fields on flat parts above the cirque, where the slush avalanches started. This water probably ran out under the snowpack and saturated lower situated localities and later triggered the slush avalanche in Malá Úpská Rokle on 9th April 1999.

Temporary brooks These brooks were fed by thawing water on the edge of the south-east facing wall of Úpská Jáma and consequently triggered a slush avalanche in April 2005 (e.g. FIGURE 3).

Roman Juras CULS, Faculty of Environment Kamýcká 129 Prague, CZECH REPUBLIC tel.: +420 605 783 195 e-mail: juras@fzp.czu.cz

Vegetation can also indicate potential starting zones of slush avalanches. Especially Sphagnum sp., Allium sibiricum, Swertia perennis and Aconitum plicatum prefer moist and very wet localities like Malá Úpská Rokle


R. JURAS, V. SPOUSTA, M. KOCIÁNOVÁ I. ŠPATENKOVÁ, J. PAVLÁSEK CZECH UNIVERSITY OF LIFE SCIENCES IN PRAGUE ŠPINDLERŮV MLÝN THE KRKONOŠE MOUNTAINS NATIONAL PARK, VRCHLABÍ

ADVANCES IN AVALANCHE FORECASTING, PODBANSKÉ

and Pramenný Důl. An entrainment and subsequent succession of the vegetation are very good proof of avalanche activity. Investigation of the succession has been surveyed since 1999 (Kociánová et al. 2005).

FIGURE 2. Full depth avalanche in Pramenný Důl valley from 1.4. 2006. A visible signs of running and spring water can be seen within a starting zone (red arrows). Photo: M. Tůma

FIGURE 1. Slush avalanche in Malá Úpská Rokle, 2.5.2002. Melting water ran down after the avalanche fall. Photo: M. Tůma

Slushflows were documented in the Krkonoše Mts. by Kociánová et. Bercik (2000) and Kociánová et. Štursová (2008). These currences gave impulse to start with a detailed study of snowpack development on high flat parts of the Krkonoše Mts. on Stříbrný Hřbet ridge and Úpské Rašeliniště mire, especially during the melting period when large slush fields are formed. Slush fields represent perfect study material for better understanding of slushflow and slush avalanche development (Juras 2009). Slush layer originates on the bottom of snowpack mostly on the mire localities. The snowpack is mainly well ripe when the mire melt water starts to saturate it. The emergence of this water is related to higher temperature rather than to precipitation, as rainfall occurs rarely during spring melting period on the top of mountains. The slush field exists only several days and velocity of its movement is negligible or zero. Time period of slush fields‘ occurrence is similar to that of slush avalanches in Malá Úpsk á Rok le.

Another survey was simultaneously done in Northern Sweden in Abisko Mts. (e.g. FIGURE 4, 5) for comparison of snowpack development, because these phenomena are well known and also well described in this region. Investigation of snow development leading to ―slushy‖ condition in both regions is interesting, because both of them are located in tundra biome, but each area is influenced by different diurnal activity (considering the polar day in high latitude of Abisko (above 68° N), which is also connected with positive net radiation lasting continuously for couple of months that highly effects snow melting). Although slushflows and slush avalanches occur mainly in northern countries, their presence was documented also in Central Europe as we report in this contribution. However, not only the Krkonoše Mts. are affected by the occurrence of these phenomena, as some slush avalanches are known from Nízké Tatry Mts. (Slovakia) Demänovská Dolina (Onesti et Hestnes 1988) as well. Prediction and identification of slush avalanches are dependent on a good knowledge of the terrain and classification of typical slush sediments (e.g. FIGURE 4, 5).


R. JURAS, V. SPOUSTA, M. KOCIÁNOVÁ I. ŠPATENKOVÁ, J. PAVLÁSEK CZECH UNIVERSITY OF LIFE SCIENCES IN PRAGUE ŠPINDLERŮV MLÝN THE KRKONOŠE MOUNTAINS NATIONAL PARK, VRCHLABÍ

ADVANCES IN AVALANCHE FORECASTING, PODBANSKÉ

FIGURE 3. Slush avalanche on the avalanche path No. 4. Sněhova Strţ in Úpská Jáma cirque. Two temporary brooks were probably the reason for its triggering. April 2005. Photo: J. Vaněk. Figure 5. Slushflow deposits from Kärkevagge. Photo: R. Juras

FIGURE 4. Small slushflow deposit in Kärkevagge, northern Sweden. 31.5.2010. Photo: M.Kociánová.

REFERENCES De Quervain, M.R., De Crecy, L., LaChapelle, E.R., Losev, K. and Shoda, M., 1973. Avalanche classification. Hydrological Sciences Bulletin 18 (4), pp. 391-402. Hestnes, E., 1985. A contribution to the prediction of slush avalanches. Annals of Glaciology 6, pp. 1-4. Hestnes, E., 1998. Slushflow hazard – Where, why and when? 25 years of experience with slushflow consulting and research. Annals of Glaciology 26, pp. 370–376. Hestnes, E. and Sandersen, F., 1998. Slushflow hazard control. A review of mitigative measures. NGI Publication 203, pp. 140–147. Juras, R., 2009. Nebezpečí břečkotoků na území KRNAP a moţnosti jejich predikce. Diploma thesis. Czech University of Life Sciences in Prague, Prague, 123 pp. (in Czech)

Kociánová, M. and Bercik, P., 2000. Břečkotok v Krkonoších (Slushflow in the Krkonoše Mts.) Krkonoše 5, pp. 6-7. (in Czech) Kociánová, M., Špatenková, I., Tondrová, A., Dvořák, I. J. and Pilous, V., 2005. Základové a smíšené laviny ve vztahu k přemísťování svahovin a dynamice vegetace. (Ground and mixed avalanches with relation to transfer of debris and dynamic of the vegetation cover). Opera Corcontica 41/1, pp. 86-99. (in Czech) Kociánová, M. and Štursová, H., 2008. Phenomena connected with thawing of snow cover in tundra zone in the Krkonoše Mts. Opera Corcontica 45, pp. 13–34. (in Czech) Perov, V. F., 1998. Slushflows: Basic properties and spreadings. NGI Publication 203, pp. 203 – 209. Onesti, J. O., 1987. Slushflow release mechanism: A first approximation. Avalanche Formation, Movement and Effects (Proceedings of the Davos Symposium, September 1986). IAHS Publ. no. 162, pp. 331-336. Onesti, J. O. and Hestnes, E., 1988. Slushflow questionnaire. NGI Publication 58200-6, pp. 112. Rapp, A., 1960. Recent developement of mountain slopes in Karkevagge and surroundings,northern Scandinavia. Geografiska Annaler 45, pp. 73– 200. Spusta,V., Spusta, V. (jun.) and Kocianova, M., 2006: Lavinový katastr české části Krkonoš v zimním období 2003/04-2005/06. (Avalanche cadaster of the Czech part of the Giant Mts. in winter season 2003/04-2005/06). Opera Corcontica 43, pp. 81–93. (in Czech)


P. ŠŤASTNÁ, M. KOCIÁNOVÁ DEPARTMENT OF NATURE PROTECTION THE KRKONOSE MTS. NATIONAL PARK

ADVANCES IN AVALANCHE FORECASTING, PODBANSKÉ

Avalanches as a natural factor – influence on the species biodiversity and the landscape P. Šťastná*, M. Kociánová Dep. of nature protection, The Krkonose Mts. National Park Administration, Vrchlabí, Czech Republic

INTRODUCTION Snow avalanches are often perceived only as a dangerous factor for the tourists and inhabitants or their properties. But the significance of avalanche activity in nature ecosystems is wider and for many organisms positive or even irreplaceable. The avalanche paths present unique biotopes because the avalanches keep these places permanently light and open in naturally forested zones. The distributed snow cover also creates mosaic of dry and moist places with specifically low form of the vegetation. For example, the probability of avalanche damage to conifers is related to the size of the tree, slow growth rates and result in small trees that can survive many years in avalanche tracks, contributing to the persistence of the avalanche community. Generally, avalanches have a relevant effect on the region's vegetation mosaic, increasing the area's community diversity and creating a fragmented vegetation pattern, while the structure and persistence of communities within avalanche tracks is connected with the frequenc y of avalanche occurrence (Erschbamer 1989, Patten and Knight 1994). Therefore avalanches provide an interesting and rich study material for many kinds of studies. THE LOCALITY The Krkonoše Mts. are the highest Mts. of the Czech Republic (300–1620 m a. s .l.) and together with the Jeseníky Mts. are the only regions with the avalanche activity in the country. A part of the mountains is also situated in Poland. Through the small area (in total: 631 km 2) the avalanche activity in the mountains is rather high – together with the polish part presents 105 avalanche paths. These avalanche paths are not influenced by the human activity because they do not threaten any villages or chalets (Štursa 2003, Spusta et al. 2006). ________________________ * Corresponding author address: Petra Šťastná Oddelenie ochrany přírody, Správa KRNAP Dobrovského 3 54301 Vrchlabí, CZECH REPUBLIC tel: +420 499456219,+420 607605941 e-mail: pstastna@krnap.cz

The leeward sides of the cirques provide also warmer stands and together with the wind flows bringing non-mountainous species from wider distances, host the unusual combination of plant species for this mountain region. These botanically rich localities are called the ―gardens― can host several hundreds of plant and moss species, tens of lichen species and certainly high numbers of insect species. For these reasons are the avalanche processes in the Krkonoše Mts. appreciated and protected. These localities present the most valuable parts of the national park at all (Jeník and Kosinová-Kučerová 1964, Šourek 1969). EXAMPLES OF CASE STUDIES CONNECTED TO AVALANCHES: The avalanche database Since the winter 1961/62 have been all avalanche paths and avalanche activity in the Krkonoše Mts. regularly monitored by the Valerian Spusta and his colleagues. The data from this long term monitoring are available in journal Opera Corcontica (Spusta and Vrba 1975, Spusta and Vrba 1991, Kociánová end Spusta 1998, Spusta et al. 2003, Spusta et al. 2006) or online at http://opera.krnap.cz/. This year will be also finished the online GIS portal for publishing of these data by Suk (2011). Dendrochronology For the prolonging of the time period of the monitored avalanche events was used a method dendrochronology. This method is based on the data gained from the trees situated in avalanche paths, because from the tree ring anomaly is possible to recognize the avalanche event. This event can be even precisely dated. The analyses of the collected data are still in process therefore no results can be presented here. Using GIS methods In this study were GIS tools used for the analyses of the slope predisposition to snow avalanche activity from the morphological point and also from the influence of the vegetation cover. To this analysis was also added with the calculation of the possible tear zones of the avalanches (Suk 2011). Similar study using statistical-probability method WoFe in GIS made also Blahůt (2008) in his model. Blahůt in


P. ŠŤASTNÁ, M. KOCIÁNOVÁ DEPARTMENT OF NATURE PROTECTION THE KRKONOSE MTS. NATIONAL PARK

his previous study (Blahůt 2007) also focused at the sorting of the avalanche types due to meteorological conditions. The botanical, lichenological, bryological and fycological surveys Because of the high inaccessibility of these localities in the vegetation season were these localities for the threaten danger of the injury often neglected by scientists in the past. Today with the rope techniques and modern equipment are these localities slowly investigated. Several new, extinct or missing species from the lichens or moss were in three last years found, many of valuable localities with rare or threaten species were found within plants, moss and lichens e.g. in the Labský důl valley (Halda et al. 2010). SUMMARY The case studies show that the avalanches are highly studied not only in a scientific way. The results would by useful to use in avalanche forecasting or for other practical purposes. REFERENCES Blahůt, J., 2007. Typy lavin Labského dolu v Krkonoších a meteorologické podmínky jejich vzniku. Opera Corcontica, 20, pp. 197–204. Blahůt, J., 2008. Mapa náchylnosti terénu Krkonoš ke vzniku lavin vytvořená pomocí nástrojů GIS a statisticko-pravděpodobnostních metod. Opera Corcontica, pp. 35–44. Erschbamer, B., 1989. Vegetation on avalanche paths in the Alps. Plant Ecology, 80/2, pp. 139– 146. Halda, J., Hauer, T., Kociánová, M., Mülhsteinová, R., Řeháková, K. and Šťastná, P., 2011. Biodiverzita cévnatých rostlin, lišejníků, sinic a řas na skalách s ledopády v Labském dole. Opera Corcontica, 48, pp. 45–68. Jeník, J. and Kosinová-Kučerová, J., 1964. Příspěvek k poznání přírody Labského dolu v Krkonoších. Opera Corcontica, 1, pp. 71–88. Spusta, V., Brzeziński, A., Kořízek, V. and Kociánová, M., 2006. Laviny v Krkonoších. Správa KRNAP, Vrchlabí, 32 pp. Kociánová, M. and Spusta, V., 1998. Lavinový katastr české části Krkonoš v období 1961/62– 1997/98. Opera Corcontica, 35, pp 3–205. Patten, R.S., Knight, D.H., 1994. Snow avalanches and vegetation patter in cascade canyon, Grand Teton National Park, Wyoming, USA. Arctic and Alpine Research, 26/1, 35–41 pp. Spusta, V., Spusta., V. and Kociánová, M., 2003. Lavinový katastr a zimní situace na hřebenu české části Krkonoš v období 1998/99–2002/03. Opera Corcontica, 40, 5–86. Spusta, V., Spusta., V. and Kociánová, M., 2006. Lavinový katastr české části Krkonoš v zimním období 2003/04 aţ 2005/06. Opera Corcontica, 43, 81–93.

ADVANCES IN AVALANCHE FORECASTING, PODBANSKÉ

Spusta, V. and Vrba, M., 1975. Lavinový katastr Krkonoš. Opera Corcontica, 12, 65–90. Spusta, V. and Vrba, M., 1991. Lavinový katastr Krkonoš. Opera Corcontica, 28, 47–58. Suk, P., 2011. Možnosti využití GIS v problematice sněhových lavin. PhD. Thesis, depon. in Lesnická a dřevařská fakulta, Mendelova univerzita v Brně, 156 pp. Šourek, J., 1969. Květena Krkonoš. Academia, Praha, 451 pp. Štursa, J., 2003. Encyclopedia Corcontica. Správa KRNAP, Vrchlabí, pp. 88.


V. ROMEO, M. FAZZINI ITALIAN STATE FOREST SERVICE, ROME UNIVERSITY OF FERRARA – DEPARTMENT OF PHYSICS AND EARTH SCIENCE, FERRARA

ADVANCES IN AVALANCHE FORECASTING, PODBANSKÉ

Winter season 2011 - 2012 in Italy: METEOMONT activity, snow and avalanche report and snow emergency on February 2012 V. Romeo1,*, M. Fazzini2 Italian State Forest Service, Rome, Italy 2 University of Ferrara – Department of Physics and Earth Science, Ferrara, Italy 1

ABSTRACT The MeteoMont National Service of the Italian State Forest Service, in collaboration with the Ministry of Defence, the Italian Air Force Meteorological Service and the Civil Protection Department, presents, for the first time in Europe, snow and avalanche report – winter 2011-2012. The history, the organization, the activities, the products and the provisions of services will be illustrated in detail. Monitoring and forecasting (Fig. 1), prevention and divulgation (Fig. 2), cooperation and technological innovations lead the service to improve safety conditions in mountainous area and not only.

altitude, below the 800 – 1000 meters (Bisci C. et al, in press; Cardillo A. et al, 2012).

FIGURE 2. The MeteoMont website information and data (www.meteomont.org) are available on mobile and i-phone applications too.

FIGURE 1. WEGO (Weather Environment GIS Oriented) is a weather forecasts, products and information system, multi-layered, that introduce a new concept of presenting web-based (by Meteorological National Service).This system is used by MeteoMont forecasters

During last winter seasons, heavy and extensive snowfalls brought several emergency situations of Civil Protection Department. These phenomena cause problems for safety due to snow and avalanche danger even at low altitude usually not affected by these precipitations. Last February in Italy, in particular in the Adriatic area (Fig. 3), the exceptional snowfall recorded the highest level of the last century. The term exceptional is statistically used for the snow cover at a low __________________________ * Corresponding author address:

Vincenzo Romeo Italian State Forest Service - METEOMONT Service Viale A. Ciamarra 139 00198 Rome, ITALY tel.: +39 06 72466252 e-mail: v.romeo@corpoforestale.it

In particular, between the end of January and mid February 2012, the Italian peninsula has been subject to the influence of many advections of arctic-continental air generating a series of cyclogenesis on the central Mediterranean Sea. This caused widely spread mainly snowy precipitation, often persistent, mostly affecting the central-southern Regions, as well as Rome and Naples and many cities in the Marche region (Fig. 4). Snowfalls and low temperatures caused some sixty casualties, besides huge damage to economic and productive facilities. The analysis of al large number of nivometric monitoring points, belonging to different institutions (MeteoMont, NeveMont, Military Aviation, Civil Protection and Universities) allowed to evaluate the spatial distribution of total snowfall. Comparing these data with those referring to past perturbation similar in length - recorded in the winter of 1929, 1956 and 1985 (Fazzini et al, 2005, Fazzini, 2007, Fazzini e Romeo, 2011) - demonstrates that the 2012 event can be considered exceptional being characterized by a higher total thickness of snow cover almost everywhere along the hills and low


V. ROMEO, M. FAZZINI ITALIAN STATE FOREST SERVICE, ROME UNIVERSITY OF FERRARA – DEPARTMENT OF PHYSICS AND EARTH SCIENCE, FERRARA

mountains of the Adriatic side of Central Italy (Fig. 5, Tab. 1).

ADVANCES IN AVALANCHE FORECASTING, PODBANSKÉ

FIGURE 3. February 14th, 2012 in Italy, in particular in the Adriatic area, the exceptional snowfall recorded the highest level of the last century

FIGURE 4. Exceptional snow cover in the city center and in the neighboring of Urbino and San Marino (Marche region).

Detailed event will be discussed, pointing out the trend, the climate meteorological development, the effects and the consequences on the interested region. It will also be showed the activities carried out by the

FIGURE .5 – Nivometric map of February 3rd, 2012 for the Italian peninsula elaborated by NeveMont data.

MeteoMont Service: the assistance to the population stranded by snow, the re-establishment


V. ROMEO, M. FAZZINI ITALIAN STATE FOREST SERVICE, ROME UNIVERSITY OF FERRARA – DEPARTMENT OF PHYSICS AND EARTH SCIENCE, FERRARA

of basic services monitoring activities, evaluation and control of the slopes from avalanches (Fig. 6). TABLE 1. Cumulated snowfall (in cm) for the most relevant events of the last century in the study area, as recorded in the available nivometric stations STATIONS URBINO CARPEGNA NOVAFELTRIA SANT'AGATA FELTRIA BAGNO DI ROMAGNA TERZO DI CARNATO FONTE AVELLANA SAN MARINO VERGHERETO PASSO SAN LEONARDO BOLOGNOLA CORTINO PETRACAMELA CERCEMAGGIORE CAPRACOTTA CAMERINO PESCASSEROLI CAMPO DI GIOVE SAN BENEDETTO IN ALPE SCANNO ATRI RIPATRANSONE TERAMO L'AQUILA CAMPOTOSTO CIVITELLA TR CAMPOBASSO CAMPO LIETO SULMONA CHIETI BOLOGNA MACERATA SERVIGLIANO PESARO VASTO RIMINI CESENATICO ROSETO

ELEV M. 473 748 293 598 471 610 691 652 872 1282 1035 1000 1000 935 1375 661 1200 1065 531 1020 444 494 300 760 1414 589 686 735 476 330 53 300 215 11 145 1 3 13

feb.'29 55 105 52 53 70 220 81 120 97 61 205 88 49 92 128 143 174 56 118 109 28 110 70 68 17 107 71 27 51 26 58 68 87 42 0 58 64 24

feb '56 111 103 61 144 115 144 125 65 137 115 151 98 113 111 295 153 155 125 115 101 93 87 67 55 146 78 99 93 82 110 63 87 108 13 15 51 56 38

jan '85

31/1-13-2/2012

44 326 63 315 50 306 85 306 60 305 105 300 63 291 73 289 63 256 130 237 95 211 - (1350 m) 24 207 (1320 m) 115 197* (1450 m) 59 190 106 184 104 175 50 170 55 165 123 161 51 155 30 145 14 145 14 145 45 144 131 138 38 138 38 136 47 130 23 129 89 105 75 100 65 88 52 86 32 83 10 81 23 59 26 50 3 47

ADVANCES IN AVALANCHE FORECASTING, PODBANSKÉ

REFERENCES Bisci, C., Fazzini M*, M. Bertrando, G. Cardillo, A. and Romeo, V.: the exceptional snowfall on February 2012 in the central Adriatic size of Italian peninsula‖ IN PRESS on Meteorologist Zeischrift Vol. 21, issue 5, 2012 Cardillo, A, Fazzini, M.,Beltrando,G.,Romeo,V., l‘enneigement exceptionnel du février 2012 dans la région de Molise (Italie centrale) in « Les climats régionaux :observation et modélisation » ACTES XXV Colloque de l‘Association International de Climatologie (Grenoble 5-8 settembre 2012), 171- 176 Fazzini M., 2007: Caratterizzazione generale dei fenomeni di innevamento in Italia. – Neve e Valanghe 60, 36–49. Fazzini M., Giuffrida, A., Frustaci,G., 2005: Snowfallanalysis over peninsular Italy in relationship to the different types of synoptic circulation: first results. – Proc. 28th Conf. on Alpine Meteorology (ICAM-2 MAP), Croatian Meteor. J. 40, 650–653. dell‘evento. – Neve e Valanghe 55, 6–15. Fazzini, M., Romeo, V., 2011: L‘enneigement dans les Apennines durant les dernier 30 ans. – Actes XXIV Colloque AIC ―Climat montagnard et risque‖, 249–254.

FIGURE 6. Large natural surface-layer slab avalanche in Pizzo Meta - Sarnano (MC) involving an important exposed transportation route, closed the day before


Advances in Avalanche Forecasting 2012  

Zborník z medzinárodnej konferencie "Pokroky v lavínových predpovediach" organizovanej pri príležitosti 40-teho výročia založenia Strediska...

Read more
Read more
Similar to
Popular now
Just for you