Risk for river flooding in lake-rich regions

The focus is to obtain discharge information on exceedance of water levels thresholds that place designated critical objects at risk for two areas in Sweden. The quantity of rain over a certain area that causes the critical discharge is also of high importance.

The case study includes the river Tabergsån (and Lake Glan) and river Storån in Jönköping County. The climate impact indicator is how often will these rivers will be flooded in a future climate compared with today’s climate and how big is the probability increasing that areas that have never flooded up to now will be flooded in the future.

The results from this case study will be used in impact-based flood forecasting.

Read Full Technical Report here!

impact based forecasting

Efforts to avoid flooding of a Swedish petrol station in Arvika year 2000

Case Study Description
Data Description
Reference information

Water-management issue to be addressed

The meandering stream Storån lies in Jönköping County in central southern Sweden, as shown in figure 1. Storån is a part of the Lagan river system that drains into Laholmsbukten which is a part of the Kattegatt Sea.

A number of polluted areas are present along the identified watercourse, which the client must manage under future climate conditions. The aim of the study is therefore to analyse how the frequency of flooding of these areas will change in a future climate, and to provide potential mitigation measures for the future.

The study consists of two parts. The first, Part A, deals with river Storån that is a meandering stream with polluted areas/sites mainly in the community of Hillerstorp. For the study a model was built in the hydraulic modelling toll HEC-RAS. The second, Part B, deals with Tabergsån and Lake Vättern. Tabergsån is a steep stream that have earlier been modelled by the author in the hydraulic modelling toll MIKE11HD. For the study the model is updated and recalibrated. There are several polluted areas/sites along Tabergsån between the lake Vederydssjön where the model starts and Lake Vättern, the end of the model. Further the tilting of lake Vättern in combination with changed discharges from the lake will be considered

The following headline messages have been identified by these studies:

  • SWICCA indicators were used to identify discharge magnitudes that will cause flooding at the polluted sites under present climate (“critical discharges”), and therefore the return period of these discharges. This was achieved using hydraulic modelling and GIS (HEC-RAS and HEC-GeoRAS, MIKE11HD and ArcMap).
  • In general, increased flows will not increase the extent or (significantly) the depth of flooding in a future climate. This result may be valid in many places across Sweden where meandering watercourses flow through fluvio-glacial catchments.
  • Critical discharges are expected to occur more regularly in the future, increasing pollutant transport downstream at risk areas. The present day critical flow is expected to occur with a recurrence interval 7.3 years by 2080, compared with 25 years in the current climate.
  • The higher land uplift in the north end of Lake Vättern will cause a shore displacement in the city of Jönköping. With changes due to climate change the combined effect of uplift and climate change will be studied

Decision support to client

During floods today the toxic chemicals can be moved to a safe level since they are inside an industrial building. The result of the case study will thus not lead to any immediate undertakings for the client. However, changes to the recurrence of critical flows means that any decision to increase the size of the industrial site in the future must be taken carefully.  

Temporal and spatial Scale

The time period most relevant to the client would not be more than 10 to 20 years. The rapid development of modelling techniques, measuring techniques and data mining could lead to the results from the case study becoming outdated after say 15 years. If it was found that the critical discharge corresponds to a 20-year flood for present time, then there is not much use to consider what will happen with discharges in a future climate. However, how rain storms will change in the future is still of interest, as heavy rains can cause local flooding.

The spatial scale is regional. The area of Jönköping County is 10 495 km².

Knowledge Brokering

Communication with the client took place via phone calls and e-mails. There was a meeting in person at the start-up of the project when the client’s wishes were discussed. During the meeting the client promised to show a list containing polluted sites and bathymetry of the stream. Bathymetric data was more difficult to supply and for that reason the more time consuming iterative method was applied, as discussed earlier.

At the start-up meeting, the challenges associated with modelling flood plains generated under fluvio-glacial conditions were also considered – namely that inundated area often does not increase with increased flow. This was confirmed by modelling and therefore only two present day and one future climate simulations were required. Recurrence interval of the critical flow was also discussed and it was emphasised that this is a more critical issue for these types of catchments.

Climate Impact Indicators

Pan-European Indicators

Indicators used are:

  • River flow (daily, seasonality and mean), extracted from SWICCA at catchment resolution and for reference period and 2080 (RCP 8.5, mean ensemble range).
  • Flood recurrence for the RCP 8.5 scenario, calculated separately using methods described in the recent progress report.
Local indicators

The local indicator of interest is changes in the return period for a specific discharge at a specific location in rivers Tabergsån and Storån, and change of rain intensity over Jönköping County.

Pan-European data to local scale

Read Full Technical Report here!


Hydrology in a future climate - Extract indicator on projected future changes of high-flow return periods (climate indicator of SIS water management) from ensemble of pan-European hydrological models. Indicators that would be interesting here are how the discharges will change in the future. Will the critical discharges return more frequently? Will they last longer?

Identification – Identify areas and objects at risk and determine the water level at which they are flooded. One also need to know how flooding will affect different polluted areas in terms of the duration of the flooding, the frequency of flooding and regarding erosion. Regarding erosion the velocity of the water would be interesting.

Data collection and model building – Collect in-data for the model. This is information about topography, information about structures in the streams, information about roughness in the stream channel and on the flood plain. Search for calibration data. The Climate Indicators will be needed here for the construction of one or several time series of discharges. The duration of high discharges is important, not least the duration of peak discharges.

Hydraulic modelling – Obtain by iteration a discharge that is consistent with the threshold water levels.

Critical rain events – Identify the rain events that cause the critical discharges. Also identify rains that could cause flooding without any contact with a water course. Here the Climate Indicators will indicate if the critical rainstorms will recur more frequently, will the last longer or is there no considerable change compared with today.

Plan of measures – If necessary make a plan to move and protect infrastructure or polluted soil that is at risk.

Lessons learnt

A number of lessons have been learnt during this work. These are summarised below:

  • The SWICCA indicators are readily accessible via the SWICCA Graphs and Download service, but require some prior knowledge in order to use them appropriately and effectively.
  • There are numerous sites in Hillerstorp containing toxic chemicals, shown in figure 3, but only one of the sites was found to be at risk of flooding in both present and future climates (2080, RCP 8.5).
  • In general, increased flows will not increase the extent or (significantly) the depth of flooding in a future climate. This result may be valid in many places across Sweden where meandering watercourses flow through fluvio-glacial catchments.
  • Critical discharges are expected to occur more regularly in the future, increasing pollutant transport downstream of at risk areas. The current critical flow is expected to occur with a recurrence interval 7.3 years by 2080, compared with 25 years in the current climate.
  • Data for the model building is easily obtained from the Swedish land survey. The most important information is the elevation data which is of high quality and accuracy.
  • Due to the lack of bathymetric data an alternative method for designing the geometry under the water surface was used. This method means that no exact roughness data is necessary for the design of the channel, but is needed for the flood plain.
  • In a case study like this most of the work lies in the construction of the hydraulic model. Since model building is very time consuming it is important first to evaluate the accuracy needed for the model. To be able to complete the work within the timeframe more effort was laid on the part of the model covering the study area Hillerstorp and less effort on the part downstream down to the downstream boundary.
  • There are numerous sites along Tabergsån containing toxic chemicals, shown in figure 2. An iteration process will be used to identify the discharges that will affect these sites. The process is to use different discharges as input in the hydraulic model MIKE11HD to obtain corresponding water levels.

Importance and Relevance of Adaptation

Polluted soil is a problem in Sweden. There are 80,000 places in Sweden where polluted soil is suspected. Many sites in the area around the steam Storån are classed as risk sites. Increased recurrence of high flows will clearly increase the risk for pollutant transport to downstream areas around the study site. These future negative effected could be mitigated by protecting or removing the polluted soil and by relocating industrial sites.  

However, to remove polluted soil is a time consuming and very costly undertaking. Only to remove the most polluted soils in Sweden would take 40 years and would cost approximately 45 billion Swedish crowns. Clearly however, this issue is of great importance if a non-toxic environment is to be handed over to coming generations.

The methods used commonly used in these types of studies are flood mapping based on specific flood return periods (i.e. the 100 year flood). The drawback in these studies is that it cannot be identified when, and at which flow rate, a specific area or object will be flooded. Previous studies have suffered from too coarse elevation data, lack of knowledge about the river bed topography, too large spacing of cross sections and one dimensional modelling. Some of these disadvantages still remain.

Earlier studies did not take climate change into account but recently, climate change is included in the calculations. However, it has been shown that an increase in discharge for the 100 year flood by 20% to 25% will not increase the flooded area significantly. Even the difference in flooded area between a 50 year flood in todays’ climate and a 100 year flood in 2098 is small. The reason for this might be of geomorphological nature rather than an increase in cross section area with an increased discharge an increase in velocity of flow takes place.

The new approach in this study is to apply return periods of different magnitudes for flood risk mapping, find out at which discharge level certain areas and objects will be flooded, and how often compared this will occur in comparison to today. These results will add more useful information to help with risk reduction.

Pros and Cons or Cost-Benefit analysis of climate adaptation

Before a cost benefit analysis can be completed, the increase of the magnitude of risk should be known, and comparisons of todays risk with the future risk should be undertaken. This is difficult to assess before obtaining model results.  Large changes in maximum discharges in the streams due to climate change are not expected, thus it is unlikely many new preventative measures and adaptations will be put in place to combat climate change overall. Rather, it is expected that preventive measures will be put in place in response to highlighting the problem that more frequent and more intense flash flooding will occur.

Policy aspects 

The Swedish parliament has set-up 16 environmental goals. One of these is to achieve a non-toxic environment. Work on contaminated sites is driven mainly by appropriation issued by the government and by Country board instruction as well as by the Environmental Protection Agency and the Water Authorities.

The County Administrative appropriations (2015, paragraph 52) states that the county administrative boards must report which measures are taken to increase the number of privately funded treatments of contaminated areas and work to remedy contaminated sites by government grants. The county administrative boards will also be coordinated and in cooperation with the Environmental Protection Agency to develop and present actions to minimize the extent of unused grants for remediation areas.

The SWICCA indicators can contribute to identify polluted areas that would be flooded and flooded more frequently in a future climate. However, in this study it was noted that only one site would be at risk in the Hillerstorp survey area.


Gustav Carlsson




Relevant EU policy


Purveyor: Gustav Carlsson, SMHI


Client: Måns Lindell, Jönköping County Administration

River Lagan (thick blue line) and Storån (thin blue line) and the study area, Hillerstorp (red box).

Site in Hillerstorp that will be affected by severe flooding in future