Suspended particulate matter (SPM) in the brackish Ebinur Lake of arid northwest China profoundly affect its water quality and watershed habitat quality. However, the actual driving mechanisms of the Lake's SPM changes remain unclear. Therefore, the purpose of this study is to explore the controlling factors driving the variability of SPM in the Ebinur Lake. This study constructed month-by-month SPM maps of Ebinur Lake based on time-series remote-sensing imageries and SPM inversion model. Thirty-four factors that might influence SPM changes were extracted, and the Partial Least Squares Structural Equation Modeling (PLS-SEM), suitable for complex relationships and factor interactions, was applied to identify the relative influence of each factor quantitatively. The results showed: (1) a clear increasing trend of SPM concentration in Ebinur Lake from 2011 to 2020; (2) that SPM changes were influenced by external and internal factors, explaining 48.2 % and 46.9 % of the changes, respectively; (3) that, to the external factors, meteorological factors exerted the greatest influence on SPM (relative contribution of 38.9 %); that, to the internal factors, water salinity imposed the greatest influence on SPM (relative contribution of 43.3 %); (4) that, among the meteorological factors, the measured variable Alashankou wind speed expressed the most significant positive effect on SPM (weighting coefficient of 0.894), and sulfate generated the strongest positive effect on SPM (weighting coefficient of 0.791) among the water salinity factors. Hence, the quantitative identification of drivers of SPM changes in Ebinur Lake could provide a new perspective to investigate the driving mechanisms of lake water quality in arid areas and inform their sustainable restoration and management.
Understanding the main driving factors of oasis river nutrients in arid areas is important to identify the sources of water pollution and protect water resources. Twenty-seven sub-watersheds were selected in the lower oasis irrigated agricultural reaches of the Kaidu River watershed in arid Northwest China, divided into the site, riparian, and catchment buffer zones. Data on four sets of explanatory variables (topographic, soil, meteorological elements, and land use types) were collected. The relationships between explanatory variables and response variables (total phosphorus, TP and total nitrogen, TN) were analyzed by redundancy analysis (RDA). Partial least squares structural equation modeling (PLS-SEM) was used to quantify the relationship between explanatory as well as response variables and fit the path relationship among factors. The results showed that there were significant differences in the TP and TN concentrations at each sampling point. The catchment buffer exhibited the best explanatory power of the relationship between explanatory and response variables based on PLS-SEM. The effects of various land use types, meteorological elements (ME), soil, and topography in the catchment buffer were responsible for 54.3% of TP changes and for 68.5% of TN changes. Land use types, ME and soil were the main factors driving TP and TN changes, accounting for 95.56% and 94.84% of the total effects, respectively. The study provides a reference for river nutrients management in arid oases with irrigated agriculture and a scientific and targeted basis to mitigate water pollution and eutrophication of rivers in arid lands.