The researchers studied soil moisture and salinity in a wetland in China using satellite data and field measurements. They created models to predict soil moisture and salinity, finding that the NDWI index was highly correlated with moisture and the D2 index was highly correlated with salinity. The BP neural network model was most accurate for estimating salinity, while a cubic function model worked best for moisture. Soil salinity was highest in the middle of the area and lowest in vegetation-covered areas, while soil moisture decreased outward from the center of the wetland.