Abstract: Quite a few indices for agricultural drought monitoring based on remote sensing technology have been developed, but their sensitivity may be affected by specific environment. Different agricultural drought monitoring indices derived from remote sensing have different temporal and spatial adaptability. For the purpose of assessing the impact of drought timely and accurately, it is very important to select appropriate monitoring indices for specific regions and specific crop growth stages. Referring to previous studies, in this paper, agricultural drought monitoring indices were divided into three types: precipitation-based, soil-based and crop-based indicators. With the relative soil moisture (RSM) as the reference, the performances of 10 drought monitoring indices were analyzed during crop growing season in Northeast China. In the process of quantitative analysis of these drought indices’ applicability, the Pearson correlation analysis was carried out on 8-day scale in the typical drought year 2009. The results showed that: (1) except in the early stage of the growing season, the absolute value of correlation coefficient between the temperature vegetation drought index (TVDI) and RSM was about 0.50 respectively. TVDI was sensitive to soil moisture, and can be used for agricultural drought monitoring in the middle and late stage of the growing season. (2) The accumulative crop water stress index (ACWSI) based on potential evapotranspiration and actual evapotranspiration was one of the indices with good correlation with soil moisture. Especially in the early and late growing season, it performed well: the absolute value of the correlation coefficient between ACWSI with RSM was above 0.47. The time scale of cumulative effect needs to be paid attention to in the application. (3) The apparent thermal inertia (ATI) was more suitable for drought monitoring in early growing season, and the modified energy index (MEI) was appropriate for various vegetation cover conditions, but it had certain instability. (4) Compared with the precipitation condition index (PCI), the accumulative precipitation condition index (APCI), considering the accumulated precipitation, reflected the soil moisture status better, especially in the middle and late stage of the growing season. So, it could be used as a supplement to other monitoring indices. (4) The vegetation conditional index (VCI) and normalized difference water index (NDWI) had low correlation with RSM, indicating low sensitivity to current soil moisture, so they are not suitable for agricultural drought monitoring in Northeast China. This study can provide some reference for the index selection of regional agricultural drought monitoring in short time scale, and construct a feasible framework for the extensive application of agricultural drought indices.

Key words: Agricultural drought, Remote sensing monitoring, Relative soil moisture, Adaptability, Northeast China