应用气象学报 (Jan 2020)
Remote Sensing Inversion of Leaf and Canopy Water Content in Different Growth Stages of Summer Maize
Abstract
Hyperspectral remote sensing technology is an important method for crop water monitoring, aiming to understand crop growth status. In order to achieve rapid, refined and comprehensive monitoring for the leaf and canopy water content of summer maize in different growth stages, controlled experiments are implemented during different growth stages of the summer maize with different irrigation water drought simulation test in North China. The water content of vegetation index (WI), water stress index (MSI), global vegetation moisture index (GVMI), compound ratio index (WNV and WCG) and reflectance curve area (Darea) of summer maize are defined for inversion models of equivalent water thickness for canopy (EWTC) and fuel moisture content for leaf (FMC). The hyperspectral remote sensing inversion models of moisture content of summer maize in 2014 are verified by using drought simulation data of different irrigation water amount during different growth periods in 2015. Results show that WI, MSI, GVMI, WNV, WCG and Darea for inversion EWTC of summer maize at the three-leaf stage doesn't pass the significance test of 0.05 level, but all the indices estimation EWTC models after the three-leaf stage pass the significance test of 0.01 level. The model accuracy for different stages from high to low are as follows: Tasseling stage, knotting stage, filling stage, mature stage, and seven-leaf stage. FMC at the seven-leaf stage and jointing stage is retrieved by 6 special indicators and all of them pass the significance test of 0.01 level. FMC at the three-leaf stage is retrieved by WNV index, but 6 spectral indicators after jointing stage cannot retrieve FMC of summer maize. In summary, the difference of precision of the same spectral indicator to retrieve the water content of summer maize is obvious in different growth stages. The retrieved water content precision is higher for middle summer maize growth period, but relatively lower for early and late remote sensing. Although canopy and leaf scale water content indices can reflect the drought situation of summer maize, considering the precision of spectral indicator retrieval of two scale water content indices of summer maize is closely related to the growth period of summer maize, a retrieval model of water content is proposed for different growth stages of summer maize to provide accurate simulation of water content in summer maiz growth.
Keywords
