Remote sensing monitoring models of soil salinization based on NDVI-SI feature space

Based on vegetation and soil information, remote sensing model was established for extracting soil salinization. In this paper, the Delta oasis of Yutian in the south of Tarim Basin was chosen as the case study for its long-term research accumulation. The relationship between salinization index (SI) and normalized difference vegetation index (NDVI) was analyzed synthetically. Through experimental analysis and theoretical derivation, the research proposed a conception of NDVI-SI feature space and discussed its biophysical characteristics. Analysis revealed that location could be used to improve the current strategies for salinization in the NDVI-SI feature space. The research presented a model of salinization detection index (SDI) to monitor severity of salinization. The results showed that the correlation between the SI and soil salt content had a higher accuracy about 0.8596. The SDI values of non-salinized soil, slightly salinized soil, moderately salinized soil and severely salinized soil were 0.399, 0.763, 0.974 and 1.201, respectively. Results of difference matrix indicated that the SDI had a higher separability in difference scale of soil salinization. The SDI demonstrates a much better performance in measuring salinity soil since it takes into account both topsoil spectrum and halophytic vegetation growth condition in the modeling process. The SDI has the potential to provide a simple and low-cost monitoring tool for assessment of salt-affected areas.