Downscaling climate change impacts on socio-economic parameters in a storyline-based investigation

Climate change has drastic impact on our lives and many socio-economic parameters. The exact consequences are hard to quantify and in general not obvious and not clear to the general public. In this work we use the storyline scenarios from within the SCENIC project for the prediction of various socio-economic parameters in warmer climates. We use machine learning algorithms to investigate how parameters that have direct impacts on society and the population develop in different warmer world scenarios. For this purpose we train echo state networks. Input data are temperature and humidity fields over Europe. Output data are time series of, e.g., mortality rates, forest fires, crop yield, and consumer price indices from Germany. The output data are predicted under the investigated climate scenarios of +2K and +4K with respect to pre-industrial time.