The effect of climate on droughts and their propagation in different parts of the hydrological cycle

Abstract

Drought is a natural hazard, caused by a lack of water availability, which can have a large impact on environment and economy. An arid area or wilting agricultural crops are in most cases the first association people have with drought, but droughts occur all over the world. The aim of this research is to gain a better understanding of the properties of droughts occurring in different climates and different parts of the hydrological cycle (precipitation, soil moisture, discharge). This is accomplished by investigating relationships between drought characteristics in different hydrological variables, climate type, and physical catchment structure. Daily time series of precipitation, soil moisture, and subsurface discharge were available on a 0.5 degree grid. These time series originate from WATCH forcing data and model outcome of a simple, spatially lumped conceptual hydrological model that was forced with these WATCH data. Time series of 1495 randomly selected cells were retrieved. Each of these cells was classified to a certain climate type (major and sub-climate) based on the Köppen-Geiger climate classification and analysed according the that type. On the time series of the selected cells, drought analyses were carried out. The drought characteristics taken into account were number of droughts, duration, and standardized deficit volume. For these drought characteristics, summary statistics and properties of 90% bi-variate probability density fields (orientation, shape, and similarity) were investigated on both major and sub-climate level. This was done for precipitation, soil moisture, and subsurface discharge. The outcomes show differences in summary statistics and shape and orientation of density fields on major and sub-climate level. These differences are related to differences in climate type and physical catchment structure. Clear differences also occurred between summary statistics and density fields of droughts in precipitation, soil moisture, and subsurface discharge, reflecting drought propagation. The observed differences helped to better explain differences in drought generation between sub-climates within a major climate type. Another aspect that was investigated in this research was the effect of using a relatively simple model, which was developed in previous research to obtain time series of different hydrological variables. The results show that the single-reservoir model is able to simulate time series relatively accurate, especially in low-flow situation which are of more importance in this drought research than peak-flows. This research concludes that drought characteristics and properties of 90% probability density fields can be related to the differences in hydrological variables (e.g. in precipitation, subsurface discharge), climate type, and physical catchment structure. This information helps drought research and management on their way to better understand drought generating mechanisms and spatial drought distribution.