CROSS-BORDER SMALL SCALE FLOOD MAPPING FOR PFRA OF THE TANA RIVER IN NORWAY AND FINLAND

Transcription

1 Working Group F on Floods Thematic Workshop on Implementation of the Directive 2007/60/EC Brno, May 2009 UNDERSTANDING OF POTENTIAL SIGNIFICANT FLOOD RISK CROSS-BORDER SMALL SCALE FLOOD MAPPING FOR PFRA OF THE TANA RIVER IN NORWAY AND FINLAND MIKKO HUOKUNA, MIKKO SANE firstname.lastname@environment.fi Finnish Environment Institute IVAR PEEREBOOM iope@nve.no The Norwegian Water Resources and Energy Directorate NIINA KARJALAINEN firstname.lastname@environment.fi Lapland Regional Environment Centre Abstract Norway and Finland are compiling preliminary flood risk assessment together for the Tana River basin. Tana River is a border river between Norway and Finland. The area of the river basin is about km 2 of which 31 % is located in Finland and 69 % in Norway. Tana River and its tributaries are in natural state and the river is well known for its salmon fishery. The river is flowing from South to North and because of that ice jams are formed quite often in the river causing flooding. However, there is only a very limited amount of property located near the river and severe flood damages are very unusual. For the preliminary flood risk assessment (PFRA), both countries are going to collect the data about the past floods. In addition to observation data, the PFRA will be based on extensive use of GISdata. Both countries use a method in which the possible flood prone areas are modelled by using a digital elevation model. Although the methods in both countries are different, both identify areas that can be dangerous which can be further used in the PFRA. The similarity of results between methods has been encouraging. For the purpose of the preliminary flood risk assessment, risk indicators within these potentially flooded areas are considered to evaluate the potential risk. These indicators, for example the number of houses and the number of inhabitants, will be based on the existing GI-data. In Finland a risk square methodology has been developed to objectively asses the flood risk. In Norway a similar approach is under development. In this paper, the methods to be used in the PFRA co-operation between Finland and Norway for the Tana River are described, with a focus on assessing if different methods can be used across the border to produce one consistent flood risk analysis. 1 (9)

2 1 Introduction The responsible authorities for the preliminary flood risk assessment (PFRA) of the Tana River are Lapland Regional Environment Centre in Finland and Narvik Regional Offices of the Norwegian Water Resources and Energy Directorate (NVE) in Norway. Based on the PFRA these authorities will make suggestions of the areas for which they conclude that potential significant flood risks exist or might be considered likely to occur (significant flood risk areas). In Finland, according to the new unconfirmed flood risk legislation the local authority makes the suggestions of the significant flood risk areas for the Ministry of Agriculture and Forestry. The Ministry makes the final decision of the naming the significant flood risk areas. In Norway, however, the national authority, NVE, makes the suggestions of the significant flood risk areas for the Ministry of Petroleum and Energy. As in Finland, the Ministry makes the final decision of the naming the significant flood risk areas. Particularly because of the cross-border river, it s important to make the selection of the significant flood risk areas in the consistent way in the both countries. NVE main office and Finnish Environment Institute (SYKE) will produce GIS-analyse for the regional authorities as a tool for identifying potential flood risk areas. By using GIS-analyse different kind of flood risk indicators of the possible significant flood risk areas will be provided for the decision-makers. The PRFA in the Tana River basin is made first in English, so the combining of Norwegian and Finnish information is easier. When the report is ready it will be translated to Finnish and Norwegian. Hearing of the parties will be arranged too. Effective flood prevention and mitigation requires coordination between Member States. The aim is the produce one single international flood risk management plan, or a set of flood risk management plans coordinated at the level of the international river basin district (the Floods Directive article 8). According to the Directive Member States shall ensure that exchange of relevant information takes place between the competent authorities concerned (article 4). In order to reach this goal Norway and Finland has established a planning group for the implementing of the Floods Directive. The Finnish members of the group are from the Lapland Regional Environment Centre and from SYKE. The Norwegian members are from NVE s regional office in Narvik and from the main office in Oslo. University of Turku is maybe going to participate in the planning group too because of the ongoing research projects related to floods in the Tana River. The first co-operation meeting concerning the Floods Directive between Finland and Norway was hold in February During the meeting the practice and the means of risk assessment were compared between the countries. 2 The Tana River Catchment The watershed of Tana is situated in the northern part of Lapland province in Finland and in the middle part of Finnmark province in Norway. The Finland's area is divided to the areas of two municipalities, Inari and Utsjoki. In Norway the watershed is divided to the areas of three municipalities, Koutokeino, Kaarasjoki and Tana. The Tana River forms together with its head waters 288 km national frontier between Finland and Norway. The river flows from south to north to the Tana fiord and the last 50 km it flows in Norway. The river is one of the Europe's most significant rivers of migratory fishes. The river including its tributary is in natural state. 2 (9)

3 Over 90 % of the watershed's area in Finland is forest and marsh. Constructed areas are situated in Utsjoki, Nuorgam and Karigasniemi. The amount of agricultural area is low. In the end of year 2006 there lived 1330 persons in Finland's side of the watershed. In Norway the biggest population centers are Tana, Kaarasjok and Koutokeino There are 14 Natura sites in the watershed which cover approximately 78 % of the watershed's area in the Finnish side. Figure 1. Tana river basin. 3 Historical floods In the Tana River watershed there are floods at some level almost every year by the time of ice break-up. Floods usually don't damage buildings, because of the morphology of the main river and the distance of the buildings from the shoreline. Roads and bridges suffer the biggest damages, when high discharges erode the soil. There is also erosion in the smaller tributaries, because in steep hillsides the small amount of water reaches fast high flow velocity. (Ollila et al. 2000) The biggest flood measured in Polmak in the lower reach occurred in 1920, when the discharge was 3843 m 3 /s. The peak discharge occurs usually around May turning to June. There are no regulated lakes or rivers in the Tana watershed. The lake percentage is 3.1 %. Because of the large size of the watershed, the hydrological events are slow excluding the floods caused by ice jams. (Pöyry 2006, Alaraudanjoki et al. 2001) Many ice jams were formed in 1966 and the flood was exceptionally large. In autumn 1998 there was an ice break-up, which formed an ice jam covering the channel from top to bottom. In spring 1999, the ice jam caused a major flood to the delta of Utsjoki. 3 (9)

4 4 The effect of climate change and long term developments Hydrological models have been used for climate change studies in Finland and Norway. According to climate change forecasts low discharges are decreasing slightly within next 70 years. The flood peak occurs earlier than nowadays and the high peak discharges are slightly decreasing. It seems that discharges in late autumns and early springs are rising. (Vehviläinen 2008) The river basin is sparsely populated and with the aid from planning, the settled areas are guided far enough from the shoreline because of the erosion. In addition, the shores of the river are quite high, which means that buildings do not get flooded easily. In the Finnish side, the Lapland regional Environment Centre determines the lowest elevations for buildings in Tana valley. The allowed base elevation for the buildings is normally the water level of a 50-year flood plus one meter. 5 Using GIS-data in flood risk estimation 5.1 PFRA the Finnish Way As flood damages in Finland have been less common than in many other parts of Europe, the lack of spatial information on major floods substantially hampers carrying out flood risk assessment. Therefore, it has been considered necessary to develop a procedure for identifying potential flood risk areas by using extensively existing GIS-data. (Sane & Huokuna 2008) GIS-analysis developed in SYKE can be used as a tool when estimating low-lying, probably flood prone areas. It also enables applying congruent criteria with identification of flood risk areas. The estimation of flood areas is based on method, in which the topography, the area of overhead catchment, lake percentage and the gradient of the river channel are taken into account. The calculation is performed by catchments and the calibrations are done with an estimated 1000-year flood discharge and water level information. The biggest problem in method is the poor accuracy of the national digital elevation model (25 x 25 m DEM). The vertical accuracy is only 1.8 m. A new 10 pixel size DEM (accuracy ~1 m) covers at this time about 70% of the country. A national laser scanning project was started 2008 to produce better DEM but it will take several years before the new DEM (2 x 2 m, accuracy better than 30 cm) is available for the whole country. The main steps in the calculation of the flood prone areas are: The pre-processing of the digital elevation model (filling the sinks and burning the river system if necessary) Modelling the flow direction network, catchments and lake percentage from the DEM Calibration of the discharge calculations (e.g. frequency analysis, flood information system) Calculation of discharge using Kaitera s nomogram Calibration of water level calculation (e.g. frequency analysis, flood information system) Calculation of water level using Bernoulli s and Manning s equations The generation of flood prone areas using the algorithm applied from "path distance" The flood prone area can be calculated for different flood frequencies. However, for the ongoing flood risk assessment work it has been decided to use a 1000-year flood as a base flood. Flood prone areas are generated first for this flood. More frequent floods can be derived from the year flood later by decreasing the flood level with a certain amount. In the PFRA-analyses the visualization of the flood risk to people and buildings will be made by using the flood risk squares. The flood prone area is divided into 250 x 250 metres squares. The 4 (9)

5 classification of the flood risk in a square is based on the amount of the people living in the square and the total floor area of the buildings in the square. The amount of the population and the floor area can be retrieved from the building and apartment registry. The classification is presented in the table 1. The squares, which have a high damage potential, are included in class I and the squares with a low damage potential are included in risk class IV. The flood risk squares gives only damage potential of people and buildings. Other possible flood damage receptors should be studied together with the flood risk squares. In the figure 2 an example of the use of the flood risk squares in the Tana River is presented. A more detailed description of the methodology has been introduced in Sane & Huokuna Figure 2. Flood prone area and flood risk squares (1/1000a) in Tana watershed according to Finnish model. Flood risk squares have been calculated only to Finnish side of the watershed. The locations of the single risk objects haven t been cross-checked. 5 (9)

6 Table 1. Classification of flood risk squares by population and floor area. Risk class Inhabitants / risk-square Floor area / risk-square I > 250 inhabitants or > m 2 II inhabitants or m 2 III inhabitants or m 2 IV < 10 inhabitants and < 250 m 2 The possible significant flood risk areas are defined by visual study of the flood risk squares and other flood damage receptors. By using the GIS-data the flood risk indicators for the possible flood risk areas are calculated. These indicators are for example the amount of population affected by flood, the number of buildings and the total floor area of the buildings in the flood prone area, number and size of special objects like hospitals, kindergartens etc. The flood risk indicators will then be used by the authorities in making the suggestions and ministry in making the final decisions of the significant flood risk areas. 5.2 PFRA the Norwegian Way In Norway the basic idea was to first develop a simple method to calculate the potential maximum rise of water levels in various kinds of rivers. Then use these maximum water level rises to determine the flood water level and interpolating these to a flood plane. Combining this flood plane with the Digital Terrain Model (DTM) makes it possible to find the potentially inundated areas. Deriving maximum rise of water levels Statistical analyses were done to derive the maximum water level rise. The method is based on the assumption that the water level can be derived without the use of detailed hydrological or hydraulic calculations. Data was used from gauging stations and from approx. 150 river stretches in Norway where hydraulic calculations have been made to produce detailed flood inundation maps. The rise of water levels from these rivers was correlated with discharge and catchment characteristics. Interpolating a flood plane Using these equations a water level rise can be calculated. A method has been developed to use the water levels to interpolate a flood plane. In regular floodplain analysis the rise in water level has been established through detailed hydraulic analysis using measured cross sections. By placing these values on the cross sections a floodplain can be calculated (interpolated). By overlaying the floodplain with a digital terrain model (DTM) the inundated area can be calculated. A method was developed to simulate cross sections using drainage pattern calculations on a virtual DTM based on buffer distance. See figures 3 and 4. The result is a set of narrow strokes perpendicular to the river. The water levels in the river can be transferred to these strokes by adding the maximum rise of water level results in the calculated flood level. 6 (9)

7 Figure 3. The 25*25 m DTM. Figure 4. The virtual DTM based on buffers and the cross section strokes. Calculating inundated areas Combining the flood plane with the Digital Terrain Model (DTM) it is possible to find the potentially inundated areas. By calculating separately the flood plane areas for river stretches with within the same 1 m maximum water level rise interval two fundamental problems were solved. 1) Water levels at the mouth of a tributary are not always dependent on runoff and field characteristics from its catchment area; they can be the result of inundating water from the main river. 2) The water divide of the virtual DTM lies beyond the actual water divide, resulting in false inundation planes. Combining the results from the different river stretches using the maximum values where results overlap will insure correct values where rivers with different water levels meet. The false inundation planes can be identified because they are not connected to the river stretch and can be removed. Risk analysis For flood risk analyses a same kind of risk square method will be used than in Finland, although a different method of weighting will be used. Preliminary tests on economic values have been done but the inhabitant database has not been available yet. These preliminary tests result in risk areas that are consistent with the locations, based on expert judgement, of detailed flood hazard maps from the national flood hazard mapping program. 6 Comparison of the generated flood areas In Norway and Finland different kind of methods are used to generate rough scale flood prone areas for the purpose of the preliminary flood risk assessment. Both methods are still under development and the Tana river analyses provided a excellent possibility to compare the results. For that purpose SYKE has used the digital elevation model from Norway's side too. The 10 x10 m DEM from the Finnish side (accuracy about +/- 1 m) was combined with the Norwegian DEM. Different coordinate systems of the DEMs were successfully combined (YKJ and UTM33). The 7 (9)

8 differences between the height systems haven't been taken into account (the Norwegians have NN54 height system and the Finns have N60 height system) The DEM was supposed to be in N60. There is an important difference in methods used to generate the flood prone area. Norwegian model takes also smaller sub-basins into account, whereas the Finnish model uses only the subbasins larger than 20 km 2. The differences in the results were mainly caused by this difference in the methods. Otherwise the results were very similar. This comparison gave a lot of information for the purpose to develop both methods. Figure 5. Comparing methods for small scale flood mapping for PFRA between Norway and Finland. A random example. Left side: Finnish analyse (light blue), right side: both Finnish and Norwegian (dark blue) analysis. The differences are mainly caused by the different method to handle small sub-basins. 7 Conclusions The preliminary flood risk assessment for the Tana River Basin will be done in co-operation with Norway and Finland. Because the both counties are going to use a same kind of method to identify flood prone areas for the PFRA, this river basin is providing a good opportunity to compare the results. There are some differences but the likeness of the results generated by both methods, especially for the larger sub basins, makes for increased credibility of the usefulness. Furthermore the results show that with different approaches chosen on different sides of the border the overall result is nonetheless consistent. Getting a consistent cross border risk assessment might prove to be more difficult because of the difference in chosen risk receptors or more important how they weigh in the process. The final selection of the significant flood risk areas will be also a political issue. A consistent method to produce information about risk receptors and indicators for the decision makers is important. Especially in sparsely populated countries were there are relatively short hydrological records, a GIS-based method to generate flood prone areas is useful. The co-operation between Norway and Finland in the Tana River basin is improving the methods in the both countries. 8 (9)

9 References Alaraudanjoki, T., Elster, M., Fergus, T., Hoseth, K. A., Moen, K., Rönkä, E. & Smith-Meyer, S Tenojoen eroosio Tenojoen säilyttäminen luonnonmukaisena lohijokena. Raportti A. Eroosio ja sedimentin kulkeutuminen (Report of soil erosion and sedimentation of the Tana River, in Finnish). Lapin ympäristökeskus, Finnmarkin lääninhallitus, NVE. Ollila, M., Virta, H. & Hyvärinen, V Suurtulvaselvitys (Report on extreme floods; assessment of the damage caused by a potential extreme flood in Finland, in Finnish, documentation page in English) Suomen ympäristö 441. Luonto ja luonnonvarat. Suomen ympäristökeskus. 138 s. ISBN ISSN Pöyry Tenonlaakson kehittämissuunnitelma ja rantaosayleiskaavat. Karigasniemen osayleiskaava. Ehdotus (Proposal for the component master plan for Karigasniemi, in Finnish). Utsjoki. Sane, M. & Huokuna, M Procedure For Identifying Automatically Possible Flood Risk Areas. EU, Working Group F, Thematic Workshop on Flood Mapping, Dublin, September, Conference paper. ation_exchange/documents_information/workshop_ Vehviläinen, B Vesivarojen muutokset, riskit ja mahdollisuudet Lapissa (Presentation of the changes, risks and possibilities of the water resources in Lapland, in Finnish). Suomen ympäristökeskus. Esitys Lapin liiton Lapin aluekehitys- ja kuntapäivillä Levi. 9 (9)