A Crop water stress index based on Remote Sensing methods for monitoring drought in an Arid area

Agricultural drought monitoring plays a vital role in ensuring food production, rational allocation and utilization of water resources, and maintaining stable and rapid economic development. The remote sensing method enables quick and non-destructively assessment of agriculture drought through crop moisture stress index (CWSI). However, existing CWSI methods suffer from limited applicability and complex calculations. To address these issues, this study introduces a new approach, CWSI-NM, which estimates CWSI by using a nonparametric approach (NP) to estimate actual evapotranspiration and the MOD16 Penman-Monteith (P-M) algorithm to estimate potential evapotranspiration. Applying Landsat 8 images and CLDAS datasets, the study found that CWSI-NM provides a clearer spatial representation of water stress compared to CWSI and TVDI. Furthermore, the correlation coefficient between CWSI-NM and ground-observed soil water content is 0.82, higher than that of traditional CWSI (0.69) and TVDI (0.65), indicating that CWSI-NM is more suitable for describing soil water stress. The results also indicate a strong correlation between CWSI-NM and ground observed soil moisture at cropland (the correlation coefficient of cropland, forest, and grassland sites was 0.91, 0.86, and 0.85, respectively), and CWSI-NM track soil water much better at cropland with a significant negative variation with observed soil water. Thus, CWSI-NM is a reliable method for long-term drought analysis, particularly for assessing spatial-temporal changes and related factors of agricultural drought.