Characteristics of Drought Propagation and Effects of Water Resources on Vegetation in the Karst Area of Southwest China.
Ying Liu,Fuzhen Shan,Hui Yue,Xu Wang
DOI: https://doi.org/10.1016/j.scitotenv.2023.164663
2023-01-01
Abstract:The topography is complex in the southwest karst region of China, with severe surface water scarcity but abundant groundwater resources. Studying drought propagation and vegetation demand for water is important to effectively protect the ecological environment and improve the management of water resources. We employed CRU precipitation data, GLDAS, and GRACE data to calculate SPI (Standardized precipitation index), SSI (Standardized soil moisture index), SRI (Standardized runoff index), and GDI (Groundwater drought index), characterizing meteorological, agricultural, surface water and groundwater droughts, respectively. The Pearson correlation coefficient was adopted to study the propagation time of these four types of droughts. The random forest method was used to identify the importance of precipitation, 0-10 cm soil water, 10-200 cm soil water, surface runoff, and groundwater for NDVI (Normalized difference vegetation index), SIF (Solar-induced chlorophyll fluorescence), and NIRV (Near-infrared reflectance index of vegetation) at the pixel scale. The propagation time of meteorological drought to agricultural drought and agricultural drought to groundwater drought in the karst area of southwest China was significantly reduced by 1.25 months compared with the non-karst area. The response of SIF to meteorological drought was faster than that of NDVI and NIRV. The importance of water resources for vegetation in the whole study period (2003-2020) was ranked as precipitation, soil water, groundwater, and surface runoff. The contribution of soil water and groundwater in forest was 38.66 %, while 31.66 % and 21.67 % for grassland and cropland, respectively, which indicated that the demands of soil water and groundwater in forest were greater than that of grassland and cropland. When drought occurred (2009-2010), soil water, precipitation, runoff, and groundwater were ranked in order of importance. The importance of 0-200 cm soil water was 48.67 %, 57 %, and 41 % in forest, grassland, and cropland, respectively, higher than precipitation, runoff, and groundwater, indicating that soil water was the main water resource for vegetation to cope with drought. Since the cumulative effect of drought on SIF was more obvious, SIF showed a more serious negative anomaly than NDVI and NIRV from March to July 2010. The correlation coefficients between SIF, NDVI, NIRV, and precipitation were 0.94, 0.79, and 0.89 (P < 0.001), and the correlation coefficients with groundwater were -0.27 (P < 0.001), -0.02 (P > 0.05) and -0.15 (P < 0.05), respectively. Compared to NDVI and NIRV, SIF was more sensitive to meteorological drought and groundwater drought and had great potential in drought monitoring.