FMI’s Climate Bulletin Research Letters e-shape Special Issue 2022

1/2022

FMI’S CLIMATE BULLETIN RESEARCH LETTERS

A

RC

H LE

T T

ER E

A

T

ER

1/2022

S RES

FMI’S CLIMATE BULLETIN

S RES E

CO NTENTS

RC

H LE

T

Editorial — 4 Seasonal preparedness pilot – sub-seasonal and seasonal prediction service for tyre companies — 5 Development of the weather and climate service “CRITERION” for the touristic sector in e-shape — 8 Urban resilience to extreme weather - sub-seasonal and seasonal forecasts for winter maintenance activities in Helsinki — 11 Harvester Seasons – a forestry service supporting climate smart operations — 14 Seasonal climate predictions for German cities to strengthen urban resilience to climate variability — 17 The Kemijoki Hydrological Forecast System – a service supporting sydropower production in Northern Finland — 21 Future climate projection of heat indices for Austrian major cities: strengthening urban resilience and meeting user needs — 24

FMI’S CLIMATE BULLETIN: RESEARCH LETTERS Volume 4 Issue 1 ISSN: 2341-6408 DOI: 10.35614/ISSN-23416408-IK-2022-01-RL

2

|

PUBLISHER Finnish Meteorological Institute (FMI) P.O. BOX 503 FI-00101 HELSINKI www.ilmastokatsaus.fi ilmastokatsaus@fmi.fi

FMI’S CLIMATE BULLETIN: RESEARCH LETTERS 1/2022

EDITOR IN CHIEF Hilppa Gregow

DESIGN Marko Myllyaho

EDITORIAL COMMITTEE Juha A. Karhu Anna Luomaranta Tiina Ervasti

Please mention the source when citing the content.

A DOI is available for each research letter article.

REVIEW BOARD ECRA community

© FMI

ED I TOR I A L

Monthly Updated Climate Services As the delegate of Finland to the Group on Earth Observation (GEO) and the EU Space program Copernicus committee, it has been a truly excellent experience to combine the high political programming with concrete and ambitious service development for specific end users. Climate change is clearly a first priority in the EU Space program, and in GEO, one of four priorities. EU policy impact assessments on Earth Observation have, since 2007, always given climate change adaptation the biggest economical impact. We are still in the beginning of the development of adaptation services, as end users, like municipalities and trade organizations, are only used to reading research reports now and then. For now, there is both a time and space limitation for service development.

Reports are produced years apart, and the information is regional at best.

End users will need to transition to using climate services more often to adapt well in time, if infrastructure needs to be adapted.

Climate adaptation is local, and it happens differently depending on the location and differently around a year, so climate services have to be able to address this. Downscaling global gridded information in time and space means concretely to have seasonal averages turning into daily timescales and information updates to be available monthly. Space-wise 100 km global grids need to be turned into limited area neighborhood scales with under 100 m grids. The task is challenging – the auxilliary information like orography, land cover/vegetation would be needed in the higher resolutions both in time and in space. Now we have openly available Earth Observation satellite data sets for many years that can provide this information. We can use machine learning to find the relationships between coarse and high resolution information.

Seasonal forecasts are the tool to enable the transition to more frequent climate services. They are produced monthly, looking forward half a year ahead and looking a full year ahead every 3 months. Their resolution is closer to climate predictions than to state of the art weather forecasts. The need to downscale and adjust biases is a prerequisite for a useful service, but after this realizes, daily ensemble information is available to provide probability distributions to the end users and this information can be used for very specific questions. Currently most users are transitioning from 10 day weather forecasts to using more statistical products.

Subseasonal (~42 day) forecasts twice a week are helping to bridge the gap for these users.

The e-shape pilot climate services have developed many applications from sub/seasonal data and the letters in this issue describe the use cases and the data value and production chains developed. Enjoy the read and get excited of what can be done already now – and imagine what could be done in a thousand more applications.

4

|

FMI’S CLIMATE BULLETIN: RESEARCH LETTERS 1/2022

MIKKO STRAHLENDORFF space advisor, FMI’s Principal Investigator of the EU e-shape -project

DOI: 10.35614/ISSN-2341-6408-IK-2022-02-RL Received 29 Nov. 2021, accepted 25 Jan. 2022, first online 03 Feb. 2022, published 02 Sep. 2022

Seasonal preparedness pilot – sub-seasonal and seasonal prediction service for tyre companies In winter, driving safety is strongly weather dependent. Timely installation of winter tyres is essential when temperature falls to near 0 °C and the surface of the roads becomes slippery due to ice and snow. A new user-oriented climate service providing tailored sub-seasonal and seasonal climate outlooks for winter tyre season and safety driving conditions was developed and piloted by the Finnish Meteorological Institute and Vianor Oy. ANDREA VAJDA, OTTO HYVÄRINEN, MIKA RANTANEN, ANDREAS TACK, MARKUS SILVENNOINEN Finnish Meteorological Institute

According to the legislation, winter tyres must be used in Finland from 1 November to 31 March if the weather and road surface conditions require (Traficom 2020). Thus, the time of winter tyre installation and distribution to the customers varies from year to year in conformity with the weather conditions. In this paper, we present a new user-oriented climate service providing tailored sub-seasonal to seasonal predictions for tyre companies, developed within the H2020 e-shape project, Seasonal preparedness pilot. The service was co-created with the tyre and car service chain, V ­ ianor Oy. The predictions provided via the sub-seasonal and seasonal climate outlooks for the upcoming six weeks and three months, respectively, allow tyre companies to plan the distribution of right tyre types, the seasonal management, and to provide relevant information for their customers about the time of tyre change. The climate service is currently in pilot phase, running in a fully automated operational mode.

Development of the sub-seasonal and seasonal climate outlooks.

An iterative strategy was followed in the creation of climate outlooks.

The end-user, Vianor, was engaged through dialog from the beginning of the design and development process

to harmonize their needs in terms of forecast products, visualization and delivery method. Information about the skill of forecasts and uncertainties were also discussed. A set of sub-seasonal and seasonal forecast products (Table 1) relevant for winter driving safety were selected and designed with the user. For the onset and offset of winter tyre season outlook the thresholds for slippery conditions, hazarding driving safety, were determined based on literature survey and on analysis of road accident data from Finland (Finnish Crash Data Institute, OTI) against weather variables for the period 2005–2018.

Data from the state-of-the-art prediction systems from the European Centre for Medium-Range Weather Forecasts (ECMWF), i.e., sub-seasonal forecasts from the ensemble prediction system (EPS) (ECMWF 2016) and seasonal forecasts from SEAS5 system (Johnson et al. 2019) were used in the development and operational runs of the forecast products.

The EPS data is produced for 46 days ahead twice a week, on Mondays and Thursdays, with 0.4° spatial resolution and is accessed through the ECMWF dissemination. SEAS5 system data is accessed from the Copernicus C3S Climate Data Store, where it is avail-

able with a horizontal resolution of 1° and updated on 13th of each month.

The skill of variables used in the development of outlooks was assessed and systematic errors from the raw data were reduced through bias adjustment to improve the quality of the forecasts. The variables were evaluated and calibrated using re-forecast data for the period 2000–2019 and ERA5 reanalysis (Hersbach et al.

2020) as reference for the same period. The quality of the sub-seasonal variables was successfully improved by applying the mean bias removal method for 2m temperature and the quantile mapping method for snow depth (Fig. 1). In case of seasonal forecasts, the biases of 2m temperature and snow depth variables were reduced by ensemble model output statistics (EMOS) (Gneiting et al. 2005).

All the variables were bias adjusted by the first pilot season (winter 2020– 2021) except the sub-seasonal snow depth, which was adjusted by the start of the pilot season 2021–2022.

After the post-processing of the input vari­ables, the outlooks were computed. The winter tyre season, probability of snow cover and freezing temperature outlooks describe the likelihood of expected climate conditions through begins after the new

FMI’S CLIMATE BULLETIN: RESEARCH LETTERS 1/2022

|

5

TIME-SCALE

Sub-seasonal outlook

CLIMATE OUTLOOKS

Winter tyre season

WEATHER CONDITIONS DESCRIBED

VARIABLES USED IN THE

METHODOLOGY AND

PRODUCT DEVELOPMENT

THRESHOLD APPLIED

Weekly probability of slippery road conditions during the forecast week in autumn

2 m air temperature

Tmin < 0 °C for 2 consecutive nights

Weekly probability of non-slippery road conditions during the forecast week in spring

2 m air temperature

Tmin > 0 °C for 7 consecutive days

Probability of snow cover

Weekly probability of snow cover Snow depth, snow on the ground density

Snow depth ≥ 1 cm

Snow depth

Weekly average snow depth

Snow depth, snow density

Weekly average snow depth

2 m air temperature

Tmean < 0 °C

Probability of freezing Weekly probability of freezing temperatures temperature Seasonal climate outlook

Probability of snow cover

Monthly probability of snow cover SEAS5 snow depth, on the ground snow density

Probability of freezing Monthly probability of freezing temperatures temperature

SEAS5 2 m air temperature

Snow depth ≥ 1 cm Tmean < 0 °C

TABLE 1: The climate outlooks co-designed and developed including the definitions and variables used.

FIG 1: Root mean square error calculated for sub-seasonal raw model, bias adjusted snow depth and observed snow climatology (2000–2019) for two locations. probability forecast, while the snow depth outlook predicts the absolute value of snow accumulation.

Production of sub-seasonal out6

|

looks. The newly developed sub-seasonal outlooks are produced in operational runs on the FMI servers using the improved version of the opera-

FMI’S CLIMATE BULLETIN: RESEARCH LETTERS 1/2022

tional service prototype developed in the CLIPS project (Ervasti et al. 2018) produce sub-seasonal climate outlooks (clips.fmi.fi). The production run

forecast is disseminated to FMI. The variables extracted from the dissemination stream are post-processed and the climate outlooks are generated.

The output data is visualized through weekly maps presenting the predicted weather conditions for each forecast product and disseminated to the users through an online user interface, the Ilmanet service platform (ilmanet. fi). The users can access the service using credentials. A user-friendly, easily understandable and interpretable format and easily accessible system were the main aspects when implementing the outlooks and the service. Each outlook consists of (1) an informational page including a short description of the forecast product and its definitions, information on the computation of the product and advice on how to interpret the forecast; (2) six static weekly maps presenting the predicted weather conditions either through probabilities or absolute values, and (3) animated version of the weekly maps (Fig. 2). The sub-seasonal outlooks are updated twice a week, on Tuesday and Friday. In September 2020, the first version of the service was set up and the winter tyre season outlook was released. The other outlooks were added in December 2020.

Production of seasonal climate outlooks. The seasonal climate outlooks are produced on WEkEO, a Copernicus DIAS services as soon as the updated seasonal forecast products from C3S CDS are shared on WEkEO. The forecast products are visualized through maps that are disseminated to the users through the seasonal.fmi.fi web-portal. The outlooks are updated on the 13th of each month. The pro-

FIG 2: An example of the winter tyre season outlook animation issued on 19.11.2020. duction of seasonal climate outlooks was deployed in January 2021.

The first pilot season was run during winter 2021–2022. After the first test season a preliminary qualitative evaluation of the forecast products was performed. Since the service wasn’t included yet to the workflow of the user during the first pilot season, their feedback on the forecast products will be received after the second pilot season. Based on the user’s feedback and the results of qualitative evaluation of the forecast products, the outlooks will be further improved to assure their usability.

This new climate service showcases the applicability of sub-seasonal and seasonal predictions tailored to sup-

port decision taking in transportation sector, allowing end-users to increase their preparedness and resilience to inter-annual climate variability.

Acknowledgements: The study contributes to the H2020 e-shape project (grant 820852) funded by the European Commission. The first version of the operational service prototype used in the production of sub-seasonal outlooks was developed in the SA CLIPS project (clips.fmi.fi) funded by the Academy of Finland (grant no.

303951). We thank the Finnish Crash Data Institute (OTI) for the road accident data provided.

ECMWF, 2016: IFS documentation, CY43R2, Part V: Ensemble Prediction System. 23 pp, https://www.ecmwf.int/sites/default/files/elibrary/2016/17118-part-v-ensemble-prediction-system.pdf.

Ervasti, T., H. Gregow, A. Vajda, T. K. Laurila, and A. Mäkelä, 2018: Mapping users’ expectations regarding extended-range forecasts. Adv. Sci. Res., 15, 99–106, https://doi.org/10.5194/asr-15-99-2018.

Finnish Transport and Communication Agency (TRAFICOM), 2020.: Winter tyres. Accessed 6 September 2020, https://www.traficom.fi/en/transport/road/winter-tyres.

Gneiting, T., A. E. Raftery, A. H. Westveld III, and T. Goldman, 2005: Calibrated Probabilistic Forecasting Using Ensemble Model Output Statistics and Minimum CRPS Estimation. Mon. Wea. Rev., 133, 1098–1118, https://doi.org/10.1175/MWR2904.1.

Hersbach, H., and Coauthors, 2020: The ERA5 Global Reanalysis. Q. J. R. Meteorol. Soc., 146, 1999–2049, https://doi.org/10.1002/qj.3803.

Johnson, S. J., and Coauthors, 2019: SEAS5: The new ECMWF seasonal forecast system. Geosci. Model Dev., 12, 1087–1117, https://doi.org/10.5194/ gmd-12-1087-2019.

FMI’S CLIMATE BULLETIN: RESEARCH LETTERS 1/2022

|

7

DOI: 10.35614/ISSN-2341-6408-IK-2022-03-RL Received 9 Nov. 2021, accepted 17 Feb. 2022, first online 18 Feb. 2022, published 02 Sep. 2022

Development of the weather and climate service “CRITERION” for the touristic sector in e-shape We present the methodological development steps as well as first results of the e-shape service CRITERION. The service is based on Copernicus ERA5 data for climatological information, the Copernicus C3S 50-member ensemble for the seasonal forecasting and on WRF local scale weather forecasts at 4×4 km over Greece. Detailed information is provided at 20 UNESCO cultural heritage sites in Greece in the form of simple symbols and charts. The service has been developed by the Research Center for Atmospheric Physics and Climatology at the Academy of Athens and is operational since April 2021 at https://unesco-weather.gr/.

STAVROS SOLOMOS¹, LIDA DIMITRIADOU¹, JOHN KAPSOMENAKIS¹, IOANNIS BINIETOGLOU², PANAGIOTIS NASTOS1,3, CHRISTOS ZEREFOS1,4,5,6 1 Research Centre for Atmospheric Physics and Climatology, Academy of Athens, Athens, Greece ²IAASARS, National Observatory of Athens, Athens, Greece ³Laboratory of Climatology and Atmospheric Environment, Department of Geology and Geoenvironment, National and Kapodistrian University of Athens, Athens,Greece ⁴Biomedical Research Foundation, Academy of Athens, Athens, Greece ⁵Navarino Environmental Observatory (N.E.O.), Messinia, Greece ⁶Mariolopoulos-Kanaginis Foundation for the Environmental Sciences, Athens, Greece

Introduction. Weather and climate information is necessary for decision-making for a range of tourism stakeholders. The decision to make a trip is greatly influenced by the physical and financial capabilities of the individual as well as by two important parameters: the individual’s desires that are formed based on the climatic conditions prevailing in the place where she/he resides and the characteristics of the desired destination, such as the natural landscape, cultural features and climate that make it attractive. A typical example is the preference of the British for destinations abroad with milder climates (Amelung et al. 2007). Furthermore, the generic climatic conditions which would attract tourists may differ substantially 8

|

from the actual prevailing conditions during their visit. Greece offers many options on which places to visit during a specific time schedule. Even the same site can offer a completely different experience when visited at different seasons. For these reasons we decided to develop a tailored weather and climate forecasting tool in the frame of e-shape, focusing on 20 sites in Greece that have been declared by UNESCO as World Cultural Heritage sites (Fig. 1a).

Service Description and co-design. The service is named “CRITERION” since its purpose is to provide an additional climate-relevant criterion for stakeholders in the touristic sector as well as to individual visitors for planning their vacations and trips. It is

FMI’S CLIMATE BULLETIN: RESEARCH LETTERS 1/2022

developed by the Research Center for Atmospheric Physics and Climatology at the Academy of Athens and co-designed with the Institute of Greek tourism confederation for the final format of the web page and user options. Co-design was mainly focused towards three directions: i) Identify the locations for the pilot service; ii) Define the requested timescales for the forecasts and iii) Design the web page. An iterative process was established with the end-user and through a series of discussions we came up with the idea to start from the emblematic UNESCO cultural heritage sites that are known worldwide and could provide a testbed for our developments. The requested forecasting window was three months for the

FIG 1: a) UNESCO cultural heritage sites in Greece, b) CRITERION service methodological outline.

Precipitation (mm)

Daily precipitation

Month FIG 2: a) Example of local climatological values and seasonal forecast for the Acropolis of Athens, b) example of local weather forecast for the Temple of Apollo Epicurius in Peloponnese. seasonal forecast and four days for the short-range forecasts to facilitate both long-term and daily vacation planning. The final web page structure was also co-designed to include a simple front page with minimal information and atmospheric symbols along with more advanced timeseries for the interested user.

The service is hosted at the Acad-

emy of Athens and is composed of three individual pillars as seen in Fig. 1b. First the local monthly climatology for each site is derived from the ERA5 dataset (Hersbach et al. 2020) for the period 2005–2019, including temperature, relative humidity, daily precipitation, and wind speed. These values are used to provide the prevailing climatological conditions at each site.

This information is accompanied with a short description of each monument and can be used as a first step for deciding the visiting locations. The seasonal forecast is derived from the 51 ECMWF System 5 ensemble members (Stockdale et al. 2018) at single atmospheric levels, interpolated at each site. The ensemble forecasts are used without bias correction. The seasonal

FMI’S CLIMATE BULLETIN: RESEARCH LETTERS 1/2022

|

9

forecast is provided for three months ahead and is given along with the climatological values of each site, to provide a measure of possible deviation of certain atmospheric variables from their average values (Fig. 2a).

Finally, the 4-day forecast is provided at each one of the CRITERION sites, based on WRF-ARW v4.1.3 model (Skamarock et al. 2019) at 4 × 4 km grid space over Greece. Such high resolution is required for the complex mountainous terrain of Greece in order to resolve local scale flows (e.g., downslopes, breezes, convec-

tive storms) and site-specific weather conditions (Solomos et al. 2021). The entire CRITERION process is fully automatic and is operational at the supercomputing cluster at the Academy of Athens. The development is mainly done in Bash shell scripting, Python scripting and Fortran environments.

Potential impact of the service.

CRITERION provides a simple interface that can be easily adopted by tourism operators such as touristic agents, hotels, Airbnb apartments etc.

The service is available at https://unesco-weather.gr/ and the web page

provides easy navigation and both simple and detailed information types for the interested visitor (Fig. 2a, b).

The developed modules can be easily expanded upon request to include also more sites in Greece. Application to other countries is also feasible and can be co-designed with potential end users to fulfil their specific needs.

Acknowledgements. The study contributes to e-shape project that is funded from the European Union’s Horizon 2020 research and innovation programme under grant agreement 820852.

Amelung, B., S. Nicholls and D. Viner, 2007: Implications of Global Climate Change for Tourism Flows and Seasonality. J. Travel Res., 45, 285–296, https://doi.org/10.1177%2F0047287506295937.

Hersbach, H. and Coauthors, 2020: The ERA5 global reanalysis. Quart. J. Roy. Meteor. Soc., 146, 1999–2049, https://doi.org/10.1002/qj.3803.

Skamarock, W.C. and Coauthors, 2019: A Description of the Advanced Research WRF Model Version 4. NCAR Tech. Note NCAR/TN-475+STR, 145 pp., http://dx.doi.org/10.5065/1dfh-6p97.

Solomos, S., P.T. Nastos, D. Emmanouloudis, A. Koutsouraki and C. Zerefos, 2021: A Modeling Study on the Downslope Wind of “Katevatos” in Greece and Implications for the Battle of Arachova in 1826. Atmosphere, 12, 993, https://doi.org/10.3390/atmos12080993.

Stockdale T., S. Johnson, L. Ferranti, M. Balmaseda and S. Briceag, 2018: ECMWF’s new long-range forecasting system SEAS5. ECMWF Newsletter, No. 154 – Winter 2017/18, 15-20, http://dx.doi.org/10.21957/tsb6n1.

10

|

FMI’S CLIMATE BULLETIN: RESEARCH LETTERS 1/2022

DOI: 10.35614/ISSN-2341-6408-IK-2022-04-RL Received 29 Nov. 2021, accepted 23 Mar. 2022, first online 23 Mar. 2022, published 02 Sep. 2022

Urban resilience to extreme weather - subseasonal and seasonal forecasts for winter maintenance activities in Helsinki Sub-seasonal and seasonal climate outlooks for winter street maintenance were developed and piloted by the Finnish Meteorological Institute in co-operation with the City of Helsinki.

OTTO HYVÄRINEN, ANDREA VAJDA, MIKA RANTANEN, ANDREAS TACK, MARKUS SILVENNOINEN Finnish Meteorological Institute

In Helsinki, there are on average 95 days per year with snow cover (based on the observation period 2000– 2019). However, winter conditions vary considerably from year to year.

The City of Helsinki is responsible for the maintenance of streets in Helsinki.

Winter street maintenance includes snow ploughing, the prevention of slipperiness by sanding, and street cleaning. Street cleaning includes sweeping up the sand and other grit accumulated over the winter months on the roadways, sidewalks and at public transportation stops, and is especially important in spring, after the snow season. Planning of these maintenance activities are currently done based on the short-term weather forecasts provided by the Finnish Meteorological Institute and from experience gained from the previous years. Within the H2020 e-shape project, under “S7-P2-Urban resilience to extreme weather”, a climate service providing sub-seasonal and seasonal climate outlooks tailored for winter street maintenance was developed in co-operation between the Finnish Meteorological Institute (FMI) and the City of Helsinki. Providing user-oriented sub-seasonal and seasonal forecast products tailored for win-

ter maintenance activities serves city authorities in planning activities, improving winter safety, and optimizing maintenance costs.

The co-design and technical implementation. From the beginning, the City of Helsinki was engaged in the co-design and development process.

Face-to-face and web-meetings were organized with the representatives of the Urban Environment Division of City of Helsinki. During the joint sessions, the potential offered by the sub-seasonal and seasonal forecasts for city management were discussed, forecast skill and uncertainties related to variables used in the development of outlooks were presented, and the most relevant sub-seasonal and seasonal forecast products were selected. Finally, the visualization and delivery modalities for the six-week and monthly outlooks were agreed.

Initially, the plan was to solely use seasonal-range forecasts. However, it became clear that the sub-seasonal forecast time scale is more relevant for maintenance work planning as most of the activities are scheduled on timescales of several weeks. Information on snow depth is needed also on the seasonal time scale for the preparation of snow dump sites and

haulage. In the end, the following tailored sub-seasonal and seasonal forecast products were selected in the co-design development process for testing. These outlooks (see Table 1 for detailed definitions) were defined and designed in collaboration with the users, based on the current maintenance practices and additional information from literature: • Snow accumulation (sub-seasonal and seasonal scale) • Longer periods of freezing weather during winter (sub-seasonal scale) • Longer periods of warm weather during winter (sub-seasonal scale) • Slippery conditions for pedestrians (sub-seasonal scale) • Conditions for spring street cleaning (sub-seasonal scale).

Some potential products were excluded due to the low skill of the variables, for example, forecasts of snow melt.

The forecast data used in the development of the sub-seasonal and seasonal climate outlooks are the extended-range forecast dataset provided by the ensemble prediction system of ECMWF (ECMWF 2016) and the SEAS5 long-range forecasting system of ECMWF (Johnson et al.

2019) available from Copernicus CDS. Sub-seasonal products are processed

FMI’S CLIMATE BULLETIN: RESEARCH LETTERS 1/2022

|

11

TIME-SCALE

Sub-seasonal (six-week) outlook

WEATHER CONDITIONS

VARIABLES USED IN THE PRO-

METHODOLOGY AND

DESCRIBED

DUCT DEVELOPMENT

THRESHOLD APPLIED

Freezing weather: 7 consecutive days with freezing weather conditions

2 m temperature

Tmean < 0 °C during > 90% of a 7-day period

Mild weather: 7 consecutive 2 m temperature days with warmer weather conditions

Tmean > 0 °C during > 90% of a 7-day period

Snow accumulation

Probability of weekly average snow depth

Snow depth, snow density

Weekly average snow depth

Slippery conditions for pedestrians

Weather conditions favorable for development of slippery conditions on sidewalks

2 m temperature, total precipitation, snow depth

Tmin = -2…2°C and Prec >0 mm OR Snow = 1 cm

Conditions for spring street cleaning

Non-slippery weather conditions when grit is not needed anymore on the streets

2 m temperature, snow depth

Tmin > 0 °C and snow < 1 cm

OUTLOOKS

Long periods of freezing and mild weather

TABLE 1: The winter preparedness outlooks developed for and with the City of Helsinki.

FIG 1: Examples of winter preparedness outlooks for the City of Helsinki: (a) weekly snow accumulation issued on 15 Nov 2021 on the Ilmatie portal, (b) monthly snow accumulation outlook published on the seasonal.fmi.fi portal on 17 Nov 2021. at FMI and disseminated through the Ilmatie/Ilmanet platform developed within FMI. The seasonal product, snow accumulation outlook, is processed in WEkEO, a Copernicus DIAS service and disseminated through the 12

|

seasonal.fmi.fi web portal. Examples of weekly and monthly snow depth outlooks are presented in Fig. 1. The service was deployed, and the first weekly outlooks published in December 2020. The project timeline is divid-

FMI’S CLIMATE BULLETIN: RESEARCH LETTERS 1/2022

ed into two Sprints, and Sprint 1 has been completed at the time of writing.

During Sprint 1, sub-seasonal temperatures and seasonal snow accumulation have been bias-adjusted. The bias adjustment of sub-seasonal snow

and precipitation will be explored in Sprint 2. Thus, the products are continuously improved in its usability.

The feedback on the outlooks.

The feedback from the City of Helsinki was gathered after the first pilot season 2020–2021. Outlooks for snow accumulation, and for longer periods of freezing or mild weather during winter were found the most useful. The outlook of conditions for spring street cleaning was consulted, but often the sand was removed from the street before the product indicated suitable conditions. The forecast of periods of mild weather was more useful for this task. Moreover, the snow condition threshold in the spring street cleaning product (less than 1 cm of snow)

proved to be too strict, as the sand can be removed even if some snow is present. For the user it is more important to know if new snow is expected in the near future.

The outlooks for slippery conditions were used as background information. However, as decisions to sand or not to sand are made on a daily basis, its benefit is still unclear. Moreover, it is also unclear if slippery conditions can be forecasted using the simple rule-based methodology used here or if more advanced methods should be explored in the future. The outlooks have been updated considering the feedback received from the user and will be tested during winter 2021–2022. The seasonal outlook of

snow accumulation became available late in the season, so it will be tested during Sprint 2.

The City of Helsinki also expressed a wish for the quantitative validation of outlooks. This is planned for the near future. Some early results for sub-seasonal snow depth and 2 m temperature validation have been presented in Rantanen et al. (2021) and Vajda et al. (2021). A quantitative evaluation of the sub-seasonal forecast products will be performed following the second testing period, when a larger data pool will be available.

Acknowledgements. The study contributes to the H2020 e-shape project (grant 820852) funded by the European Commission.

ECMWF, 2016: IFS documentation CY43R2 - Part V: Ensemble Prediction System. 23 pp, https://www.ecmwf.int/node/17118.

Johnson, S. J., and Coauthors, 2019: SEAS5: The new ECMWF seasonal forecast system. Geosci. Model Dev., 12, 1087–1117, https://doi.org/10.5194/ gmd-12-1087-2019.

Rantanen, M., M. Kämäräinen, O. Hyvärinen, and A. Vajda, 2021: Skill assessment of sub-seasonal 2-m temperature forecasts for Helsinki, Finland.

EMS Annual Meeting 2021, online, 6–10 Sep 2021, EMS2021-138, https://doi.org/10.5194/ems2021-138.

Vajda, A., O. Hyvärinen, M. Rantanen, A. Tack, and M. Silvennoinen, 2021: Seasonal preparedness pilot – sub-seasonal and seasonal service for tyre companies. FMI’s Clim. Bull. Res. Lett., 4(1), 5-7, https://doi.org/10.35614/ISSN-2341-6408-IK-2022-02-RL.

FMI’S CLIMATE BULLETIN: RESEARCH LETTERS 1/2022

|

13

DOI: 10.35614/ISSN-2341-6408-IK-2022-05-RL Received 14 Dec. 2021, accepted 12 Apr. 2022, first online 11 May 2022, published 02 Sep. 2022

Harvester Seasons – a forestry service supporting climate smart operations The Finnish Meteorological Institute has developed Harvester Seasons, a web-based service for the forestry sector aiming to pave the way for climate smart operations.

This service is supporting the forestry sector by offering a dedicated trafficability and soil information service based on weather and seasonal forecast models. This service leverages the European efforts on climate monitoring by linking Copernicus DIAS platforms, the Copernicus Climate Data Store, and Copernicus seasonal forecasts with climate data records.

MIRIAM KOSMALE1, JAAKKO IKONEN1, MIKKO MOISANDER1, TUOMO SMOLANDER1, HEIKKI OVASKAINEN2, ASKO POIKELA2, MIKKO STRAHLENDORFF1 1

Finnish Meteorological Institute Metsäteho Oy

2

Background. The Finnish forestry sector traditionally plays a significant role for the national economy. More than 75 % of Finland is covered with forests, which leads to the wood production being responsible. More than 75 % of Finland is covered with forests, which leads to the wood production being responsible for 1/5 of Finland’s export market. Approximately 2.7 billion euros from the forestry sector is an important part of Finland’s annual tax revenue (Finnish Forest Industries 2021). Commercial timber harvesting in Finland takes place mechanically with wheel-based forest machines.

The machines used by forest machine contractors and operators need good load-bearing conditions for their operations due to their heavy weight.

A typical harvester chain consists of a harvester and a forwarder, of which the forwarder can weigh almost 40 tons when fully loaded (Fig. 1).

According to the Finnish Forest Act on forestry, wild species, and ecosystems, the “damage to the terrain which may weaken the growing conditions for the tree stand must … be avoided” (Ministry of Agriculture and 14

|

FIG 1: Heavy and middle-sized forwarder operating on peatland.

Forestry 2014). Especially the number of deep ruts greater than 10 cm caused by heavy machinery has to be kept minimal. On surfaces with low soil bearing capacity, heavy harvesting machines can cause damage and prevent adequate regeneration of the ground.

Finnish forest can be classified into areas of good harvestability, depending on annual season, weather condition and soil condition. Based on ALS Airborne Laser Scanning via LIDAR measurements and a digital elevation model, Metsäkeskus (Finnish Forest

FMI’S CLIMATE BULLETIN: RESEARCH LETTERS 1/2022

Centre) is providing static trafficability mosaics for Finland (Metsäkeskus, 2021). Those maps are differentiating between mineral soil and peatlands and classifying whether those surfaces have sufficient soil bearing capacity. Classes are color-coded from green to red, thus representing the trafficability on different soil conditions (Fig. 2, right) and seasons. In Finland, the share of peatlands from which wood can be harvested only in winter is relatively high. Within the recent years good winters for harvesting have decreased. Warmer winter

FIG 2: Screenshot of trafficability provided by Harvester Seasons web application from location near Lappeenranta (61.25331N,27.97171E). 6 month forecast on trafficability starting from selected date is given as time series. Shortterm forecast (red in time series) is based on FMI 10-day weather forecast information.

like in 2019/2020 (Lehtonen 2021) made it necessary to adapt harvesting operation planning. The challenge in timber harvesting is to optimize the operation planning according to actual carrying capacity of the ground to avoid significant shortfalls in production. Smart operation planning does not only have monetary effect but also help protecting our forest environment and ecosystem.

Harvester Seasons (https://harvesterseasons.com) is a web application developed by the Finnish Meteorological Institute that supports the Finnish forestry sector by offering dedicated trafficability and soil information system.

Copernicus, the backbone for Harvester Seasons. Copernicus, Europe’s eyes on Earth, offers access to global and local data from satellites, groundbased measurements, or model data.

Harvester Seasons input data stream is built upon this backbone of Copernicus data.

Copernicus Climate Change Service (C3S), implemented by ECMWF,

produces monthly updates of daily seasonal weather forecasts for 6 months ahead (Copernicus 2021a) as well as grid-based reanalysis data by ERA5-Land (Muñoz Sabater 2019). By bias-adjusting this seasonal weather forecast with higher resolved climate statistics of ERA5-Land, it is possible to ingest 51 highly resolved ensemble forecasts into the Harvester Seasons service backend. Trafficability information is then derived by combining statistically model forecasts of soil parameters and seasonal forecast data with the lidar based static surface classification datasets. For the shortterm forecast FMI’s 10-day weather forecast information is used for service prediction of trafficability. In collaboration with the experts from the forestry sector and based on works of Sirén et al. (2019) and Bergqvist et al. (2018), thresholds for hydrological and meteorological conditions were set in order to derive a dynamic classification (Fig. 2, right). Even though long-term seasonal forecasts up to half a year are still challenging

(Hyvärinen et al. 2020), the approach of dynamic trafficability indices helps the forestry sector in climate smart operation planning.

WEkEO is one of the 5 EU Copernicus DIAS reference services for environmental data, virtual processing environments and skilled user support (Copernicus 2021b). WEkEO has direct access to the Climate Data Store, which leverage C3S data for Harvester Seasons. Backend and frontend of the Harvester Seasons service is completely implemented within the WEkEO virtual machines.

Co-design and user feedback.

Within the funding frameworks of Copernicus Programme’s Climate Change Service (C3S) as well as the e-shape project, the primary focus of this service is on direct user interaction. As co-design partners FMI had Metsäteho and with its customers and clients, direct access to feedback from the forestry sector. Metsäteho’s shareholders are all large Finnish forest companies (Metsä Group, Stora Enso, UPM) and state forests (Metsähallitus),

FMI’S CLIMATE BULLETIN: RESEARCH LETTERS 1/2022

|

15

for which private harvesting contractors offer harvesting services.

Service developers implemented user requirements iteratively. These requirements were, e.g., an easily understandable map layer, index-based trafficability information (good vs bad) and the necessity of direct API access for user interfaces. Color-coding for the dynamic weather depen­ dent trafficability has been established in a similar way to the already existing static trafficability maps. This helps users to easily grasp the new service information by its well-established look (Fig. 2). Foremen of harvesting machines as well as operation planners require condensed information for it to have a real added value for their daily work.

As a special request by stakeholders, selected land cover characteristics (Copernicus 2021c) like tree cover density and FMI’s forest fire index

are included in the service. Harvesting operations with machines equipped with steel tracks increases the risk for light fires within forests. In Finland authorities give warnings during high forest fire risk and harvesting operation activities should be suspended.

The forest fire warning system is designed to inform the general public and the fire authorities about forest fire risk caused by dry terrain (Vajda et al. 2014).

Summary. The Harvester Seasons service has been launched for a greater Finland domain in May 2020. The web application is focusing on trafficability to support planning of harvesting operation and fostering Earth Observation and model prediction data in business operations. Having direct access to Copernicus climate data records and having them translated to a tailored information content bridges the gap between science and

community.

Service Info. The service is accessible at https://harvesterseasons.com.

Find updated info on LinkedIn. A video on how to use the service is avail­able on Youtube.

Acknowledgements. This service was developed as a use case by the Copernicus Programme’s Climate Change Service (C3S) contract under framework agreement ECMWF/ COPERNICUS/2019/C3S_428g_FMI: https://climate.copernicus.eu/harvester-seasons. The service was partly funded by e-shape. The e-shape project has received funding from the European Union’s Horizon 2020 research and innovation programme under grant agreement 820852: https:// e-shape.eu/index.php/news-events/ harvester-seasons-new-climate-service-to-help-with-estimating-conditions-for-harvesting-timber.

Bergqvist, I., E. Willén, J. Peuhkurinen, K. Väätäinen, H. Lindeman, A. Laurén, and A. Poikela, 2017: Forest trafficability maps – data sources and methods. EFFORTE Deliverable 3.3, 15 pp, https://projects.luke.fi/efforte/wp-content/uploads/sites/14/2017/09/EFFORTE-D3.3-Forest-trafficability-maps-data-sources-and-methods.pdf.

Copernicus, 2021a: Seasonal forecasts. https://climate.copernicus.eu/seasonal-forecasts.

Copernicus, 2021b: WekEO – Copernicus and Sentinel data at your fingertips - Data and Information Access Services (DIAS), https://wekeo.eu.

Copernicus, 2021c: Land Monitoring service: Land cover characteristics; https://land.copernicus.eu.

Finnish Forest Industries, 2021: Future growth and sustainable welfare. Accessed 14 December 2021, https://www.metsateollisuus.fi/en/facts-aboutfinnish-forest-industry.

Hyvärinen O., A. Venäläinen, and A. Vajda, 2020: Bias-adjusted seasonal forecasts of soil moisture for forestry applications in Finland. Adv. Sci. Res., 17, 23-27, https://doi.org/10.5194/asr-17-23-2020.

Lehtonen, I., 2021: Record mild winter of 2019/2020 in most of Finland. FMI’s Clim. Bull. Res. Lett., 3(1), 4–7, https://doi.org/10.35614/ISSN-23416408-IK-2021-02-RL. Metsäkeskus, 2021: Korjuukelpoisuus. Accessed 14 December 2021, https://www.metsakeskus.fi/fi/avoin-metsa-ja-luontotieto/metsatietoaineistot/ korjuukelpoisuus.

Ministry of Agriculture and Forestry, 2014: Forest Act (1093/1996; amendments up to 567/2014 included), 16 pp, https://www.finlex.fi/en/laki/kaannokset/1996/en19961093_20140567.pdf.

Muñoz Sabater, J., 2019: ERA5-Land hourly data from 1981 to present. Copernicus Climate Change Service (C3S) Climate Data Store (CDS). Accessed 14 Dec 2021, http://doi.org/10.24381/cds.e2161bac.

Sirén, M., A. Salmivaara, J. Ala-Ilomäki, S. Launiainen, H. Lindeman, J. Uusitalo, R. Sutinen, and P. Hänninen, 2019: Predicting forwarder rut formation on fine-grained mineral soils. Scandinavian J. Forest Res., 34, 145-154, https://doi.org/10.1080/02827581.2018.1562567.

Vajda, A., A. Venäläinen, I. Suomi, P. Junnila, and H. Mäkelä, 2014: Assessment of forest fire danger in a boreal forest environment: description and evaluation of the operational system applied in Finland. Met. Apps., 21: 879-887. https://doi.org/10.1002/met.1425.

16

|

FMI’S CLIMATE BULLETIN: RESEARCH LETTERS 1/2022

DOI: 10.35614/ISSN-2341-6408-IK-2022-06-RL Received 26 Nov. 2021, accepted 31 May 2022, first online 02 Jun. 2022, published 02 Sep. 2022

Seasonal climate predictions for German cities to strengthen urban resilience to climate variability The Deutscher Wetterdienst (German Meteorological Service, DWD) provides operational seasonal climate predictions that are statistically downscaled for selected German cities. The new climate service, based on the German Climate Forecast System, was developed within the European Union Horizon 2020 project e-shape and is part of the DWD climate prediction website at www.dwd.de/climatepredictions.

SASKIA BUCHHOLZ, ANDREAS PAXIAN, BIRGIT MANNIG, AMELIE HOFF Deutscher Wetterdienst

Potential impact of the service. The DWD recently launched a new climate service for selected German cities based on seasonal climate predictions for the two variables, near-surface (2 m) temperature and precipitation.

The new service aims at decreasing the vulnerability of urban population to hazardous weather events and risks caused by climate variability. Climate predictions for urban areas can provide the scientific basis for midterm planning decisions. The new product is tailored for city authorities helping to plan and prepare, e.g., for periods of anomalous high temperatures, so that preventive measures can be taken in the occupational health and safety sectors. With regard to long-lasting dry periods, the decision makers can plan the distribution of water re­ sources among different sectors (e.g., agriculture, industry, and urban green irrigation) more effectively.

Description of the pilot. Seasonal predictions give a prognosis on the development of temperature and precipitation for the coming 1-6 months.

The German Climate Forecast System Version 2.1 (refined version based on GCFS2.0) (Fröhlich et al. 2021) provides the operational seasonal predictions for the new climate service. It

is based on the global coupled earth system model of the Max Planck Institute for Meteorology (MPI-ESMHR) (Müller et al. 2018, Mauritsen et al. 2018). Exact initial conditions of a forecast are crucial for the forecast skill. For the climate forecast to be as close as possible to the real state of the atmosphere, observations are assimilated into the MPI-ESM-HR prior to the start of the predictions. The current GCFS2.1 uses continuous nudging to bring the state of the model climate close to the observation-based state of the climate (Baehr et al. 2015, Fröhlich et al. 2021).

For a robust statistical estimate of the quality and reliability of the predictions, a large number of historical forecasts starting in the past (also called re-forecasts or hindcasts) are calculated for each forecast and compared to observations. Each forecast starts with slightly varying initial conditions of the climate system. The resulting variety of solutions, also called ensemble, is used to evaluate the uncertainties caused by uncertainties in the assimilated observations and the non-linearity of the climate system.

In GCFS2.1 the ensemble is generated by different methods in atmosphere and ocean. For the hindcasts,

GCFS2.1 starts 30 ensemble members for each calendar month in the years between 1990-2020. Real-time forecasts are performed with 50 ensemble members per forecast to achieve even more robust forecasts than the 30 members in the past (for which 50 members would not be feasible due to the high computational demand).

For Germany, the spatial resolution of the global climate predictions is increased using the empirical-statistical downscaling method EPISODES (Kreienkamp et al. 2019). EPISODES consists of two modules. In the first step, the statistical relationship between large-scale processes (e.g., large-scale relative humidity fields), prescribed by reanalysis data or the global climate forecast model and high-resolution precipitation and temperature observations, are defined by the combination of the selection of analogue days with a follow-up regression. In the second step, multivariate time series are created with a weather generator on the high-resolution target grid for various output variables. The HYRAS gridded data (Razafimaharo et al. 2020, Rauthe et al. 2013) are used as near-surface observational data for the downscaling.

The EPISODES output has a horizon-

FMI’S CLIMATE BULLETIN: RESEARCH LETTERS 1/2022

|

17

FIG 1: Map of the seasonal ensemble mean prediction of temperature for the e-shape pilot city Aschaffenburg. (Figure source: DWD)

18

|

FMI’S CLIMATE BULLETIN: RESEARCH LETTERS 1/2022

FIG 2: Time series of the seasonal probabilistic prediction of precipitation for the German capital city Berlin. (Figure source: DWD) tal grid point distance of about 5 km.

The seasonal prediction for each city is an average of all grid boxes within the city boundary. In the case the city area is small, fewer than the usual minimum number of 9 grid boxes are used for aggregation. This is still feasible as

the EPISODES output is no dynamical model output (which needs to be aggregated for larger areas) but is based on real observations. The new service includes maps (Fig. 1), time series (Fig. 2) and tables of the ensemble mean and probabilistic predictions in

combination with the prediction skill of temperature and precipitation for the 16 German state capital cities and Aschaffenburg (e-shape partner for the service development).

The quality (i.e., skill) of the climate prediction is evaluated against the skill

FMI’S CLIMATE BULLETIN: RESEARCH LETTERS 1/2022

|

19

of the reference prediction climate mean (climatology) observed over the evaluation period, which is commonly used by climate data users if climate predictions are not available. For the evaluation we use all available hindcasts, which encompass the initialisation years 1990–2020. The skill score of the ensemble mean prediction is determined using the skill score of the mean-squared error between the ensemble mean of the hindcasts and the actual observation (MSESS). The prediction skill score of the probabilistic predictions is determined using the ranked probability skill score (RPSS), which examines to which extent the predicted probabilities of occurrence of the categories (e.g., ‘below normal’, ‘normal’, ‘above normal’, thresholds are the 33rd and 66th terciles of the reference period 1991–2020) agree with the category actually observed.

The reference prediction for the skill score is the observed climatology. A bootstrapping method is then applied to verify whether the skill comparison between hindcasts and a reference prediction is subject to accidental variations due to small sample sizes (significance test).

EPISODES conserves the skill of the downscaled hindcasts for the two variables, near-surface (2 m) temperature and precipitation. Thus, the seasonal information is available at a higher spatial resolution without losing skill. Furthermore, the output of

the statistical downscaling is nearly bias-free, which is, beside the higher spatial resolution, an added value for the climate service (Ostermöller et al.

2021).

Co-design. The design of the DWD climate prediction website has been developed in close cooperation with users from various sectors. At the beginning of the service development for German cities, a small survey was carried out among eleven cities for their requirements on the climate prediction products. During two individual meetings with representatives from the municipality of Aschaffenburg (departments of urban planning and environmental protection), the city’s individual needs for the service (e.g., graphs and description of temperature extremes and threshold exceedances, information on hot and dry episodes and heavy precipitation) could be discussed. During the last decade, city authorities in Germany have gained knowledge to deal with climate projections, impacts and uncertainties. Despite the achieved goals in the interpretation of climate projections, there are still knowledge gaps for the vast majority of cities when it comes to climate predictions and their added value.

The website content is presented in two sections, “basic” and “expert”, to cater for the specific needs of different users. The “basic climate predictions” place a particular emphasis on

comprehensibility, while the “expert climate predictions” offer more detailed information, including prediction skill maps. City authorities attach great importance to data quality and uncertainties. To offer a low-threshold access to this information, a user-oriented traffic light shows at first glance the qualitative information about the prediction skill (“basic climate predictions”). It indicates whether regional mean climate predictions are significantly better (green), not significantly different (yellow) or significantly worse (red) than the reference prediction. A feedback formula on the website offers the possibility to express suggestions, praise, and criticism. Furthermore, the DWD holds an annual workshop for end-users and informs twice per year about new innovations and changes via a newsletter.

Acknowledgements: We thank Kristina Fröhlich, Sascha Brand and Katharina Isensee for their support with GCFS2.1, as well as Philip Lorenz and Frank Kreienkamp for their support and further adjustments of the EPISODES model. Besides, we would like to thank Klaus Pankatz for his help with the evaluation, Miriam Tivig for her support with the climate forecast website and Barbara Früh for her conceptual guidance. The e-shape project is funded under the European Union’s Horizon 2020 research and innovation programme under grant agreement 820852.

Baehr J., and Coauthors, 2015: The prediction of surface temperature in the new seasonal prediction system based on the MPI-ESM coupled climate model. Clim. Dyn., 44, 2723–2735, https://doi.org/10.1007/s00382-014-2399-7.

Fröhlich, K., and Coauthors, 2021: The German Climate Forecast System: GCFS. J. Adv. Model. Earth Syst., 13(2), https://doi.org/10.1029/2020MS002101.

Kreienkamp, F., A. Paxian, B. Früh, P. Lorenz, and C. Matulla, 2019: Evaluation of the empirical-statistical downscaling method EPISODES. Clim. Dyn., 52, 991-1026. https://doi.org/10.1007/s00382-018-4276-2.

Mauritsen, T., and Coauthors, 2018: Developments in the MPI-M earth system model version 1:2 (MPI-ESM1.2) and its response to increasing CO2. J. Adv. Model. Earth Syst., 11(4), 998-1038, https://doi.org/10.1029/2018MS001400.

Müller, W. A., and Coauthors, 2018: A Higher-resolution Version of the Max Planck Institute Earth System Model (MPI-ESM1.2-HR). J. Adv. Model.

Earth Syst., 10(7), 1383-1413, https://doi.org/10.1029/2017MS001217.

Ostermöller, J., P. Lorenz, K. Fröhlich, F. Kreienkamp, and B. Früh, 2021: Downscaling and Evaluation of Seasonal Climate Data for the European Power Sector. Atmosphere, 12(3), 304, https://doi.org/10.3390/atmos12030304.

Rauthe, M., H. Steiner, U. Riediger, A. Mazurkiewicz, and A. Gratzki, 2013: A Central European precipitation climatology – Part I: Generation and validation of a high-resolution gridded daily data set (HYRAS). Meteorologische Zeitschrift, 22(3), 235-256, https://doi.org/10.1127/09412948/2013/0436.

Razafimaharo, C., S. Krähenmann, S. Höpp, M. Rauthe, and T. Deutschländer, 2020: New high-resolution gridded dataset of daily mean, minimum, and maximum temperature and relative humidity for Central Europe (HYRAS). Theor. Appl. Climatol., 142, 1531-1553, https://doi.org/10.1007/ s00704-020-03388-w.

20

|

FMI’S CLIMATE BULLETIN: RESEARCH LETTERS 1/2022

DOI: 10.35614/ISSN-2341-6408-IK-2022-07-RL Received 17 Dec. 2021, accepted 31 May 2022, first online 03 Jun. 2022, published 02 Sep. 2022

The Kemijoki Hydrological Forecast System – a service supporting hydropower production in Northern Finland The Finnish Meteorological Institute has recently developed a web-service that pro­ vides detailed hydrometeorological information to support more efficient hydropower production operations in Northern Finland. This service aims to reduce uncertainties in freshet forecasts through assimilation of satellite data-based cryosphere products to an in-house developed hydrological modelling system (HOPS) and an in-house developed machine learning streamflow forecasting algorithm. The service leverages common European efforts in climate monitoring by linking data from the Copernicus Climate Data Store with Copernicus seasonal forecasts data records as well as Copernicus Earth Observations.

JAAKKO IKONEN1, CEMAL TANIS1, MATTI KÄMÄRÄINEN1, MIRIAM KOSMALE1, HEIKKI POIKELA2, MIKKO STRAHLENDORFF1 1

Finnish Meteorological Institute Kemijoki Oy

2

Background. Kemijoki Oy is the most important hydropower and adjustable power producer in Finland. Kemijoki Oy operates 16 hydropower plants and 4 regulated reservoirs in the Kemijoki watercourse area. These simultaneously act as the most important flood regulation structures in Northern Finland. Currently, hydropower production accounts for approximately 19% of total energy production in Finland.

Freshet inflows to hydropower reservoirs, typically from mid-April/early-May to the end of June, accounts for 55 to 70% of the total annual inflows to reservoirs in the Kemijoki watershed. It is therefore essential to have careful planning and reservoir management schemes in place well before the onset of the peak inflow volumes. Reservoir management is important as the energy demand is out of phase with the natural availability of the water resources; typically, demand is higher during the colder months when the inflows are lower

and vice versa. Hydropower producers need to redistribute the availability of these resources from the spring inflow periods to other times of the year when electricity demand is higher, i.e., during the six months of colder weather while maintaining the balance between a sufficiently large volume of water for optimal production and enough remaining capacity for safe flood risk management.

Traditional hydrological forecasting in the Kemijoki watershed. FMI’s Hydrological Operations and Predictions System (HOPS) model is used operationally by FMI to provide daily hydrological nowcasts, 10-day deterministic forecasts and 90-day ensemble forecasts for Kemijoki Oy over the entire Kemijoki watershed (FMI 2021).

The HOPS modelling system consists of a distributed and modified version of the Sacramento Soil Moisture Accounting Model (SAC-SMA) (Burnash 1995) coupled with a modified version of the SNOW-17 temperature

index snow accumulation and ablation model, a soil temperature model based on Rankinen et al. (2004) and an in-house developed distributed routing model accounting for overland, and channel flow retention and attenuation, based on hydrographic properties. The parameters for the SAC-SMA model are derived from a semi-physical a-priori parametrization scheme introduced by Koren et al. (2000). Satisfactory simulation results have been achieved (with an average Nash-Sutcliffe coefficient (NS) of efficiency of 0.79 and an average Pearson correlation of 0.94 between observations and hindcast simulations across selected validation points, see e.g., Fig. 1.) by running the modelling system without any calibration apart from manually adjusting parameters governing snowmelt rates.

Machine learning for a novel approach for hydrological forecasts.

Machine learning algorithms have also been tested and implemented in the

FMI’S CLIMATE BULLETIN: RESEARCH LETTERS 1/2022

|

21

FIG 1: (a) Location of the Kemijoki HOPS modelling domain (red dots: regulated flow points, blue dots: natural flow observation points) and (b) Streamflow simulation (HOPS) timeseries at the Marraskoski observation point and comparison with Finnish Environment Institute’s observations (OBS).

Kemijoki watershed with very promis­ ing results. The machine learning model used is based on gradient boosting. The machine learning process consists of constructing a carefully designed predictor set to which an optimized gradient boosting regression model between the predictor data and the streamflow observations in each measurement station using historical data between 2007–2020 is fitted. As predictors, the model uses the ECMWF HRES deterministic temperature and precipitation forecasts over the watersheds of measurement stations, the analytical representations of annual and daily cycles, and the observed streamflow of each station in an autoregressive manner. To be able to simulate the cumulative effect of, e.g., precipitation and melting processes, the weather model output is lagged in time multiple times; for example, when predicting the streamflow at t=0, precipitation values of the catchment grid cells from t=0, t=-1, t=2, ... t=-10 time steps are considered.

The model was proven to be able to produce good forecasts despite a relatively small and restricted set of predictors, and with a more complete 22

|

predictor data set, including, e.g., satellite observations and predicted snow depths, even better performance is expected.

Assimilation of satellite-based cryosphere products. Assimilation of Earth Observation (EO) based snow observations can significantly reduce uncertainties in hydrological model snowpack simulation. Currently Copernicus EO-based Snow Water Equivalent GlobLand SWE (Copernicus 2021) data is operationally assimilated into the HOPS modelling procedure. The use of EO-based snow observations offer two main types of advantages: 1) EO-based snow observations are independent from point scale observations (usually used to drive hydrological models), and 2) they provide basin-wide information and data on the state of the snowpack as opposed to point scale observations.

Snowfall is a dominant control vari­ able in SWE accumulation. We modify snowfall forcing according to EObased SWE observations to obtain the required model SWE accumulation constraint. In our application, the corrections are determined by comparing

FMI’S CLIMATE BULLETIN: RESEARCH LETTERS 1/2022

EO-based SWE observations with an initial model simulation and then applying the corrections retrospectively as part of a second model simulation.

Example results of SWE assimilation can be seen in Fig 2. and Fig 3. Assimilation of EO-based SWE data to HOPS model simulation runs in, for example the Kemihaara sub-basin, ge­ nerally improves overall peak streamflow forecasting results. However, this is not always the case. In 2018, for example, the peak streamflow error to observed streamflow peak is increased from 141 m3/s to 365 m3/s.

The NS coefficient and Pearson correlation between observed and simulated streamflow for the spring peak runoff period (April 1st to June 30th) is however not significantly affected. From 2019 onwards, assimilation of EO-based SWE observations improves spring streamflow forecast results with varying success. However, in general, improvements of both Pearson correlation and NS are minimal apart from the spring of 2021, where NS is increased from 0.49 to 0.89. The more significant improvements in simulation accuracies manifest improvements in peak streamflow error while

FIG 2: Kemihaara streamflow observations (see Fig. 1a) and HOPS simulation runs with (Qsim SAT) and without (Qsim GD) SWE assimilation (left, 2018 and right, 2019).

FIG 3: Kemihaara streamflow observations (see Fig. 1a) and HOPS simulation runs with (Qsim SAT) and without (Qsim GD) SWE assimilation (left, 2020 and right, 2021). peak flow timing errors remain largely unchanged. In spring 2019, the peak streamflow error is reduced from 141 m3/s to 86 m3/s and in 2020 from 247 m3/s to 105 m3/s as well as in 2021 from 451 m3/s to only 46 m3/s. In the

spring of 2021, however, the peak flow timing error is increased from -1 days to -5 days.

Acknowledgement. The service is funded by the e-shape project. The e-shape project has received fund-

ing from the European Union’s Horizon 2020 research and innovation programme under grant agreement 820852 (e-shape 2021).

Burnash, R.J.C., 1995: The NWS river forecast system—catchment modeling. Computer Models of Watershed Hydrology, V.P. Singh, Ed., Water Resources Publications, 311–366, https://www.wrpllc.com/books/cmwhn.html.

Copernicus, 2021: Snow Water Equivalent. Copernicus Global Land Service, Accessed 17 December 2021, https://land.copernicus.eu/global/products/swe. e-shape, 2021: Pilot 7.4 | Hydropower in snow reservoir – climate service. Accessed 17 December 2021, https://e-shape.eu/index.php/showcases/ pilot7-4-hydropower-in-snow-reservoir-climate-service.

Finnish Meteorological Institute (FMI), 2021: Kemijoki Hydrological Forecasting System. Finnish Meteorological Institute, Accessed 17 December 2021, https://hops.fmi.fi/.

Koren, V.I., M. Smith, D. Wang, and Z. Zhang, 2000: Use of Soil Property Data in the Derivation of Conceptual Rainfall-Runoff Model Parameters.

Proc. 15th Conf. Hydrol. 2000, Long Beach, CA, Amer. Meteor. Soc., 103–106.

Rankinen, K., T. Karvonen, and D. Butterfield, 2004: A simple model for predicting soil temperature in snow-covered and seasonally frozen soil: model description and testing, Hydrol. Earth Syst. Sci., 8, 706–716, https://doi.org/10.5194/hess-8-706-2004.

FMI’S CLIMATE BULLETIN: RESEARCH LETTERS 1/2022

|

23

DOI: 10.35614/ISSN-2341-6408-IK-2022-08-RL Received 29 Nov. 2021, accepted 31 May 2021, first online 03 Jun. 2022, published 02 Sep. 2019

Future climate projection of heat indices for Austrian major cities: strengthening urban resilience and meeting user needs Based on Earth Observation and modelling data, climate indicators for urban regions can support sustainable urban planning, risk management and environmental protec­tion. The current study demonstrates the development of climate services for major cities in Austria by providing climate projections on an urban scale. This approach focuses on heat indices and improves the data uptake by end users through standard­ izing modelling methodology, as well as formats, and facilitating data access. This enables quick visualization using data services of the Austrian national Climate Change Data Centre.

SANDRO OSWALD1, MAJA ZUVELA-ALOISE1, MICHAEL AVIAN1, CHRIS SCHUBERT2 Zentralanstalt für Meteorologie und Geodynamik (ZAMG), Vienna, Austria Vienna University of Technology (TU Wien)

1

2

Urban heat in changing climate. The local surface energy balance is altered in cities due to (i) high fraction of sealing and (high) building density as well as (ii) comparably low fraction of vegetation land cover. As a result, urban areas have higher ambient temperatures than rural ones, also known as the Urban Heat Island (UHI) effect (Landsberg 1981; Oke 1982). Due to the rising urbanization and growing fraction of sealed surfaces in total land use, temperature is increasing in urban areas. Hence, more and more people suffer during extreme heat waves as, e.g., in the summers of 2003 (García-Herrera et al. 2010) or 2015 (Urban et al. 2017). In order to counteract the negative effects of heat waves, which are expected to intensify even further in the future (IPCC 2013), and develop urban planning strategies for the mitigation of the UHI, detailed information on thermal conditions within a city is required, including possible development of heat extremes in the future.

24

|

Urban resilience to extreme weather. The European Commission implemented several climate services, such as the Copernicus Climate Change Service (C3S), providing climate indicators based on Earth Observation (EO) information and modelling data in order to support different types of end users and decision makers tackling the problems of climate change.

In order to improve the user uptake of the EO data the H2020 project e-shape (EuroGEO Showcases: Applications Powered by Europe) aims to gather these EO resources in Europe and enhance the capabilities of existing European services in addressing different societal challenges. Furthermore, e-shape supports the coordinated exploitation of EO data and services through 27 cloud-based pilot applications in seven thematic areas, as well as the generation of information for these applications together with the users.

As part of e-shape, the showcase “Urban resilience to extreme weather”

FMI’S CLIMATE BULLETIN: RESEARCH LETTERS 1/2022

consists of three pilots with the goal to strengthen resilience and adaptive capacity to extreme weather in summer season on various timescales (e.g. from seasonal forecasts to climate projections). The pilot “Climate projection products for Austrian cities”, carried out by Austrian National Meteorological and Geophysical service (ZAMG), demonstrates the development of climate services for urban regions by creating climate projections for various Austrian cities. This is done by providing high spatial resolution climate information based on the EO data and modelling, focusing on heat indices. Data are made freely avail­ able on a national climate data platform (Climate Change Centre Austria (CCCA) data) to support urban planning, risk management and environmental protection. Thereby, the pilot follows the standardization of the method for urban climate modelling and its output to ensure a comparable modelling basis for other European cities.

FIG 1: Schematic workflow of deriving climate indices based on EO data, urban climate model simulations and regional climate projections for the major cities in Austria. Top and mid left: Modelling system including integration of the EO data and below the cuboid method is illustrated. Lower left: Number of summer days for the city of Vienna as a 30-year average for the scenarios RCP4.5 and RCP8.5 as well as the periods of 2021-2050 and 2071-2100. Lower right corner: Number of summer days for Vienna (spatial information, time series) visualised on the Climate Change Centre Austria (CCCA) data server. Download here for full-size image. (Source: ZAMG)

Urban climate model simulations.

Within the cooperation with Germany’s National Meteorological Service (DWD), the ZAMG carried out urban climate model simulations with the MUKLIMO_3 model (Sievers 2016), which allows utilization of various satel­lite products of Copernicus services (e.g. Urban Atlas, tree cover

density etc.) as input data to describe land surface distribution and properties. With the simulation of radiation, soil and atmospheric temperature, relative humidity, and wind flow in urban areas on a 3D grid (100 m spatial resolution, vertical resolution 10–100 m), the model provides the basis for evaluation of the (surface) tempera-

ture distribution within urban areas.

In combination with meteorological measurements and regional climate model scenarios (representative for the background climate by applying the cuboid method (Früh et al. 2011, Žuvela-Aloise et al. 2018, Geletič et al. 2019)), the model can be used to derive several climate indices related

FMI’S CLIMATE BULLETIN: RESEARCH LETTERS 1/2022

|

25

to heat conditions such as the annual average number of summer days (days with maximum air temperature ≥ 25 °C). The regional climate model simulations from the EURO-CORDEX project (Jacob et al. 2014) provide information on meteorological vari­ ables until 2100 under various climate change scenarios, e.g., RCP4.5 and RCP8.5 (IPCC 2013). An ensemble of model output for different RCP scenarios is used to project future heat conditions in urban areas. In the pilot, eight different combinations of bias corrected regional/global climate model output were used to calculate climate indices for each year between 2011 and 2100 (Fig. 1).

Co-design of climate service and data uptake. In order to support urban planning and decision-making processes, urban climate projections and representation of climate information have been gaining importance among the researchers and stakeholders in recent years (e.g. de Wit et al. 2020). In case of urban resilience in Austria, continuous interaction between climate scientists, city admin-

istrations and urban planning experts have pointed out several key issues, which could improve the uptake of research results: (i) data accessibility, (ii) standardization of data formats, (iii) improved data quality with description of uncertainties, and (iv) easy data presentation. To tackle at least aspects i, ii, iv and provide data as a service regarding FAIR Guiding Principles (Wilkinson et al. 2016), the Climate projection products are curated by the CCCA data repository (CCCA 2015) as part of the Austrian research data infrastructure.

The urban climate projections data are distributed in the NetCDF format and follow the standards of NetCDF Climate and Forecast (CF) Metadata Conventions (CF-MC 2010) and Open Geospatial Consortium (OGC 2022) web services. The data collection for this pilot is available under e-shape (2022). The CCCA Data Centre enables an on-the-fly interactive data visualisation of NetCDF data with its server architecture, e.g. THREDDS Data Server (Fig. 1, bottom right). The user (registration required) can adapt

and modify style parameters (e.g. colour scheme, portrayal) and the map design. More relevant is the on-demand processing. A time slider function allows to dive into different time scenarios and automatically creates a graph of the time series from 2011 until 2100 for a single spatial point.

In addition, an integrated CCCA subset service (Schubert et al. 2019) allows for the creation and persistently republishing of any data fragments, filtered by an “Area of Interest” or a specific time frame. Source code of these functionalities is provided as open source on Github (Github 2008).

Acknowledgements: We thank the European Commission H2020 project e-shape (grant agreement No. 82085) for their support. The German Meteorological Service (DWD) for providing the urban climate model MUKLIMO_3.

The Climate Change Centre Austria for access to data services. We gratefully thank the cities of Vienna, Linz, Salzburg, Innsbruck, Graz, and Klagenfurt for providing diverse geoinformation.

CCCA, 2011: Climate Change Centre Austria. Accessed 09 May 2022, https://data.ccca.ac.at.

CF-MC, 2010: Climate and Forecast (CF) Metadata Conventions. Accessed 09 May 2022, https://cfconventions.org. e-shape 2022: e-shape Project-Urban resilience to extreme weather. Accessed 09 May 2022, https://data.ccca.ac.at/group/e-shape-urban-resilience-to-extreme-weather.

Früh, B., and Coauthors, 2011: Estimation of climate-change impacts on the urban heat load using an urban climate model and regional climate projections. J. Appl. Meteor. Climatol., 50(1), 167–184, https://doi.org/10.1175/2010JAMC2377.1.

García-Herrera, R., J. Díaz, R. M. Trigo, J. Luterbacher, and E. M. Fischer, 2010: A review of the European summer heat wave of 2003. Crit. Rev. Environ. Sci. Technol., 40(4), 267-306, https://doi.org/10.1080/10643380802238137.

Geletic, J., M. Lehnert, P. Dobrovolny, and M. Zuvela-Aloise, 2019: Spatial modelling of summer climate indices based on local climate zones: expected changes in the future climate of Brno, Czech Republic. Climatic Change, 152, 487–502, https://doi.org/10.1007/s10584-018-2353-5.

Github, 2008: CCCA-DC at GitHub Inc. Accessed 09 May 2022, https://github.com/ccca-dc/ckanext-thredds.

IPCC, 2013: Climate Change 2013: The Physical Science Basis. Contribution of Working Group I to the Fifth Assessment Report of the Intergovernmental Panel on Climate Change, Cambridge University Press, Cambridge, United Kingdom and New York, NY, USA, 1535 pp, https://doi.org/10.1017/CBO9781107415324.

Jacob, D., and Coauthors, 2014: EURO-CORDEX: new high-resolution climate change projections for European impact research. Reg. Environ.

Change, 14, 563–578, https://doi.org/10.1007/s10113-013-0499-2.

Landsberg, H. E. (editor), 1981: The Urban Climate. Academic Press, 275 pp, ISBN: 9780080924199.

Oke, T. R., 1982: The energetic basis of the urban heat island. Quart. J. R. Met. Soc., 108, 1-24, https://doi.org/10.1002/qj.49710845502.

Schubert C., G. Seyerl, and K. Sack, 2019: Dynamic Data Citation Service—Subset Tool for Operational Data Management. Data, 4, 115, https://doi.org/10.3390/data4030115.

Sievers, U., 2016: Das kleinskalige Strömungsmodell MUKLIMO 3 Teil 2: Thermodynamische Erweiterungen. Berichte des Deutschen Wetterdienstes, 248, 151 pp, ISBN 978-3-88148-490-9, http://nbn-resolving.de/urn:nbn:de:101:1-201606173510 (pdf).

Urban, A., H. Hanzlíková, J. Kyselý, and E. Plavcová, 2017: Impacts of the 2015 heat waves on mortality in the Czech Republic—A comparison with previous heat waves. Int. J. Environ. Res. Public Health, 14, 1562, https://doi.org/10.3390/ijerph14121562.

Wilkinson, M., and Coauthors, 2016: The FAIR Guiding Principles for scientific data management and stewardship. Sci. Data, 3, 160018, https://doi.org/10.1038/sdata.2016.18. de Wit, R., and Coauthors, 2020: Supporting climate proof planning with CLARITY’s climate service and modelling of climate adaptation strategies–the Linz use-case. Urban Climate, 34, 100675, https://doi.org/10.1016/j.uclim.2020.100675. Žuvela-Aloise, M., K. Andre, H. Schwaiger, D. N. Bird, and H. Gallaun, 2018: Modelling reduction of urban heat load in Vienna by modifying surface properties of roofs. Theor. Appl. Climatol., 131, 1005-1018, https://doi.org/10.1007/s00704-016-2024-2.

26

|

FMI’S CLIMATE BULLETIN: RESEARCH LETTERS 1/2022

FMI’S CLIMATE BULLETIN: RESEARCH LETTERS 1/2022

|

27

Ilmatieteen laitos ilmastokatsaus@fmi.fi www.ilmastokatsaus.fi