Drought monitoring of the maize planting areas in Northeast and North China Plain

Abstract

Remote sensing technology can cost-effectively access to a wide range of real-time land surface spatial information, therefore it is widely used in agricultural crops monitoring. And timely assessment of drought stress using remote sensing is valuable for improving the crop management level. In this study, the drought stress was evaluated in the maize planting areas of the Northeast and North China Plain using satellite remote sensing technology. The absolute distance index was used to identify the spatial pattern of the maize planting areas using the moderate resolution imaging spectroradiometer (MODIS) derived enhanced vegetation index data that processed with Savitzky–Golay filter. The temperature vegetation drought index (TVDI) was derived from MODIS normalized difference vegetation index and land surface temperature data. Simultaneously, soil moisture content (SMC) was obtained for an eight-day period and matched with satellite remote sensing data to characterize the drought stress of maize. After estimating the field capacity and wilting point, the SMC was further transformed into the real available water content (RAWC) of the soil. Our results indicated that negative correlations between RAWC and TVDI was observed for the 0–10 cm (R2 = 0.594) and 20–30 cm (R2 = 0.641) soil layers, respectively, which was significantly better than the correlation between the SMC and TVDI in the 0–10 (R2 = 0.396) and 20–30 cm (R2 = 0.499) soil layers. Moreover, the sensitive period regarding the maize water demand were identified (day of year 161, 169, 233, 241, and 249) via the linear regression analysis of the measured maize yield and TVDI for eight-day intervals. Furthermore, the weights of the water demand of maize in different growth stages was determined via multiple linear regression analysis, and a comprehensive drought index (TVDISW) was established to evaluate the drought stress in the whole maize growth period. Eventually, a comprehensive analysis for the drought stress in the study area was conducted combined with various environmental factors. Furthermore, the performance of this method was independent of the meteorological and field survey data, which may facilitate the mapping of drought monitoring of crops planting areas at a large scale.