Abstract

Water conservation is important for maintaining a well-functioning ecosystem. Using the Soil & Water Assessment Tool (SWAT) model, we simulated the hydrological cycle and evaluated the spatial and temporal variability of water conservation in the Dongjiang River. Spatial autocorrelation analysis and geographically weighted regression (GWR) were used to investigate the effects of landscape patterns and normalized difference vegetation index (NDVI) on water conservation. Four hydrological stations with regular rate, validation period deterministic coefficient R2, and efficiency coefficient ENS (Nash–Sutcliffe) met the adaptation requirements, and the model accurately simulated the hydrological process of the Dongjiang River. The interannual variation of water conservation showed a non-significant upward trend (0.12 mm/10a), and the spatial distribution is given as follows: midstream > downstream > upstream. Regression coefficients of NDVI of high, medium, and low-density forest land in the sub-basin with water conservation decreased from the upper to lower reaches. The negative correlation of LPI, CONTAG, and IJI was concentrated in the upper reaches and southeast of the basin, and that of SHDI was concentrated in the middle and lower reaches; the regression coefficients of LPI and PLAND of forest land decreased from middle reaches to north and south, and ED decreased from southwest to northeast. The positive correlation of grassland was concentrated in the watershed’s upper reaches; the regression coefficient of construction land showed an increasingly negative correlation from the lower to upper reaches. This study provides scientific support for promoting the hydrological ecosystem and comprehensive utilization of water resources in the Dongjiang River basin.

Full Text
Published version (Free)

Talk to us

Join us for a 30 min session where you can share your feedback and ask us any queries you have

Schedule a call