The winning team used integrated monitoring (IM) data from SLU to check that their contributed model could simulate the soil moisture in the forest landscape in a correct way. The mentor for the winning team from Göttingen was Anneli Ågren, researcher at the Department of Forest Ecology and Management, SLU in Umeå.
In integrated monitoring (IM), a very large number of variables are measured in the environment, of which soil moisture is a part. SLU carries out the IM program together with Geological Survey of Sweden, SGU and Swedish Environmental Research Institute (IVL) on behalf of the Swedish Environmental Protection Agency. The soil moisture data are freely available online.
Collection of data on soil moisture
The data on soil moisture is collected through a combination of manual and automated measurement techniques. At present, the data is a high-resolution time-series collected from the areas of Gammtratten and Kindla. Three different types of places provide the data: at the top of a hill, on the slope, and down in a valley. The topographically different measurements provide knowledge about the movements of the water in the ground.
Measurements in time and space
From an integrated monitoring perspective, it is highly interesting to extend measurements, both in regards to the number of areas but also to obtain a long time series. A long time series allows more correct trend analyses. Trends in turn provide an opportunity to better understand the present and the future, predict processes, and put changes (such as climate change) in perspective. Stefan Löfgren, Research Group Leader at the Department of Aquatic Sciences and Assessment, SLU has extensive experience in integrated monitoring of various processes in the soil:
- We want to be able to perform good environmental analyzes of how different substances move in the soil as there is a risk that they will be transported to groundwater and surface water. Therefore, we need knowledge of the water process in the ground. Soil moisture measurements provide important information about when and how acids, bases, nutrients, trace metals, and more would be added to different types of forest water. This would be a very good indication of how to avoid unnecessarily large human impact.
Anneli emphasizes how interesting it would be to expand these measurements, since the hydrological processes are a little known but still very important, not least for the areas' biogeochemistry.
The use of data by the winning team
The winning team in the hackathon used IM's time series on soil moisture to test whether their AI model indicated correctly the soil moisture. The results showed a successful alignment between measured IM data and the AI model. In the future, the winning model could be used by forest companies to reduce road damage, which can become extensive when the ground is damp. This would mean reduced costs, both economic and environmental. The forest companies' felling and operating costs are reduced, moreover they reduce the risk of having to restore damaged areas (a statutory obligation). Reduced ground damages from driving also benefit water quality and, in the long run, biodiversity. Driving with heavy machinery in wet areas causes major ground damage and affects water quality. Methylmercury can form in standing water bodies created in the tracks, and changed flow paths for water can lead to browner water, which affects everything that lives there.
The winning entry shows how data from integrated monitoring can be used in a good way that creates several types of benefits.
