Evaluating the spatiotemporal dynamics of driving factors for multiple drought types in different climate regions of China

• Published in: Journal of hydrology (Amsterdam) Vol. 640; p. 131710
• Main Authors: Ding, Yibo, Lu, Zehua, Wu, Lingling, Zhou, Li, Ao, Tianqi, Xu, Jiatun, Wei, Renjuan
• Format: Journal Article
• Language: English
• Published: Elsevier B.V 01.08.2024
• Subjects: Driving factor; Drought; ERA5; Random Forest; ERA5; Drought; Driving factor; Random Forest
• ISSN: 0022-1694
• DOI: 10.1016/j.jhydrol.2024.131710

•PET and SM might be primary driving factors for hydrological drought in China.•Precipitation, soil moisture, and temperature generally had relatively importance to droughts.•Vegetation evapotranspiration played a key role in meteorological drought in southern China.•The importance of driving factors might be periodic dynamic changes to drought.•The authors declared that they have no conflicts of interest to this work. Extreme weather events and natural hazards are increasingly prevalent with global warming, making it imperative to focus on the driving factors and dynamic changes associated with drought—a disaster with widespread impacts and significant losses. This study used a longtime-series of ERA5 reanalysis data to elucidate the dynamic change characteristics of direct and indirect driving factors across China. Ground observation data spanning an extended period were used to estimate the performance of the reanalysis data. The selected driving factors for drought included precipitation (PRE), potential evapotranspiration, soil moisture (SM), bare soil evapotranspiration (BSET), vegetation canopy evapotranspiration (VCET), subsurface runoff (SUBRO), surface runoff (SURRO), temperature (TEM), specific humidity, and near-surface wind speed (WS). Meteorological drought, hydrological drought, and agricultural drought were delineated using the Standardized Precipitation Evapotranspiration Index, Standardized Runoff Index, and Standardized Soil Moisture Index. The Random Forest method was used to ascertain the relative importance of driving factors and their main contributions to different types of droughts. The findings revealed that ERA5 exhibited great simulation performance compared to observational data. However, the correlation coefficient of SM might be relatively lower than that of other driving factors across China. Notably, BSET and PRE exhibited higher relative importance degrees (RID) for meteorological drought in subregions 1, 2, and 3. At the same time, SUBRO demonstrated a more statistically significant influence in subregions 3, 4, and 5. SURRO played a statistically significant role in subregion 1. Moreover, BSET and PRE were pivotal for meteorological drought in northern China, whereas VCET assumes importance for hydrological drought in southern China and the North China Plain. Across different types of droughts, PRE, SM, and TEM generally exhibited relatively large RID values in China. The dynamic changes in RID for both direct and indirect driving factors varied for different drought types across different subregions of China. Notably, the RID displayed contrasting dynamic changes between PRE and BSET for meteorological drought in most subregions. Additionally, the degree of relative importance of near-surface WS tended to be lower for hydrological drought across various subregions. This study contributed to a deeper understanding of drought evolution and enhanced monitoring efforts.