Spatio-temporal analysis of remotely sensed soil moisture and vegetation patterns during recent European droughts

Drought occurrence and drought severity are likely to increase in the future due to more extreme weather conditions. Drought preparedness and mitigation can be achieved with the help of skilful drought forecasts and accurate quantifications of both drought severity and drought impact. For that to be possible, we first need to be aware of the spatiotemporal evolution of agricultural droughts and their main effects on vegetation. Therefore, in this study, we evaluated patterns in soil moisture and vegetation during the large scale droughts that occurred in between 2000 and 2018.

Soil moisture data were obtained from the CCI v04.5 dataset. ­­­Vegetation was analysed using (1) the Normalized Difference Vegetation Index (NDVI), a common approach to quantify vegetation cover, and (2) near-infrared reflectance of vegetation (NIR_v), which has been shown to strongly correlate to Gross Primary Production (GPP) and canopy development. Both vegetation datasets were derived from MODIS reflectance data. All three datasets were normalized to allow for an in-depth comparison of drought patterns in both space and time.

Correlations were found between soil moisture and vegetation data, and showed the possibility to discern the occurrence of water- and energy- limited vegetation. In both North and South Europe, soil moisture dry anomalies show lower correlation with dry anomalies in vegetation compared to central Europe: the highest correlations were consistently found in between 55-60˚N. In Northern Europe, the lower correlations are likely due to vegetation being energy limited rather than water limited, even during soil moisture drought events. In Southern Europe, on the other hand, it can be argued that the vegetation is better adapted to drought conditions and, as a result, the drought has less impact on vegetation. This analysis shows that drought impacts are not only related to drought severity, but also to latitude and thus climate.

­­In addition, we look at time lags between soil moisture droughts and drought impacts on vegetation, a comparison between the different years in drought over the past two decades and a comparison between drought signatures per precipitation regime. The results of this study will provide insights on the evolution of different droughts and their effects on vegetation. They could be used in future efforts regarding agricultural drought forecasting, because main trends can be predicted.