Comparison and evaluation of different dryness indices based on vegetation indices-land surface temperature/albedo feature space

摘要

Nowadays, there are many dryness indices based on vegetation indices (VI_s)-land surface temperature (T_s) feature space. Which dry-ness index should we use in dryness monitoring? The differences and capabilities for dryness monitoring among seven indices were evaluated in this study. These dryness indices were based on VI_s-T_s or Albedo feature space. For instance, Temperature Vegetation Dryness Index (TVDI) from a triangle NDVI (Normalized Difference Vegetation Index)-T_s feature space (TVDI_t), TVDI from bi-parabolic NDVI-T_s feature space (TVDI_c), TFDI from FPAR (Fraction of Absorbed Photosynthetically Active Radiation)-T_s feature space, TVDI from LAI (Leaf Area Index)-T_s feature space (TLDI), TVDI from EVI (Enhanced Vegetation Index)-T_s feature space (TEDI), TVDI from SAVI (Soil Adjusted Vegetation Index)-T_s feature space (TSDI) and Vegetation Condition Albedo Dryness Index (VCADI) from NDVI-Albedo feature space. In this study, the assimilated 5/10 cm depth soil moisture, field measured 10 cm depth soil moisture, precipitation, and the assimilated land surface temperature data were selected as the indicators to test the performance of each dryness index in two periods of 2013/2015 and three periods of 2018. The results showed that the spatial distributions of dryness from TVDI_c, TVDI_t, TFDI, TLDI, TEDI, and TSDI were similar, except for VCADI. The ability of TSDI and TVDI_t in dryness monitoring was not as good as TVDI_c and TEDI, however, the robustness of them was stable and can be an alternative dryness index. TFDI can be applied to evaluate dryness conditions, but its robustness was not stable and its monitoring performance was not as good as other indices in Shaanxi province, China. Although TLDI can detect dryness conditions when NDVI reached saturation, its robustness was worse than TEDI. When NDVI did not reach saturation, TVDI_c had the best ability in dryness monitoring. TEDI was the optimal dryness index when NDVI was approaching or reaching saturation. This study determined that NDVI reached saturation (LAI = 3) when its average value was >0.5368.