Heatwaves are extreme near-surface temperature events that can have substantial impacts on ecosystems and society. Early Warning Systems help to reduce these impacts by helping communities prepare for hazardous climate-related events. However, state-of-the-art prediction systems display low accuracy with forecasts of heatwaves more than two weeks in advance, which are required for advance warnings. In this article and accompanying analysis, AER scientist Judah Cohen collaborated with European colleagues to develop machine learning methods to understand and predict central European summer heatwaves on timescales of several weeks. First, we identify the most important regional atmospheric and surface predictors based on previous studies and supported by a correlation analysis: two-meter air temperature, 500-hPa geopotential heights, precipitation and soil moisture anomalies in central Europe, as well as Mediterranean and North Atlantic sea surface temperatures, and the North Atlantic jet stream. Based on these predictors, we apply machine learning methods to forecast two targets: summer temperature anomalies and the probability of heatwaves for 1–6 weeks lead time at weekly resolution. For each of these two target variables, we use both a linear and a random forest model. The machine learning models outperform persistence and climatology at all lead times. For lead times longer than two weeks, the machine learning models compete with the ensemble mean of the European Centre for Medium-Range Weather Forecasts’ hindcast system. We thus show that machine learning can help improve sub-seasonal forecasts of summer temperature anomalies and heatwaves.
Figure 1. Lagged linear correlations between the predictors and the temperature in the extended summer season (MJJAS) at weekly time resolution. Hatched cells correspond to non-significant linear Pearson correlation coefficients at 5% significance level.
Citation: Sub-seasonal Prediction of Central European Summer Heatwaves with Linear and Random Forest Machine Learning Models
Weirich-Benet, M. Pyrina, B. J. Estevez, E. Fraenkel, J. Cohen, & D. Domeisen
Artificial Intelligence for the Earth Systems, 2023