Remote sensing of vegetation drought:Research progress
Drought has been a serious threat to Chinese food and ecological security.Satellite remote sensing has unique technical advantages in vegetation drought monitoring and early warning,and it is an important means to further improve the defense ability of drought disaster in most countries including China.Based on the perspective of drought disaster system,this paper firstly divided the remote sensing monitoring methods of vegetation drought into three categories:drought-causing factor monitoring method,vegetation condition monitoring method,and comprehensive monitoring method.For the first category,the degree of drought was expressed mainly by measuring the degree of abnormality of the drought-causing factor,which usually utilizes a long time series of drought-causing factor data(i.e.,rainfall,soil moisture,air temperature,evapotranspiration,etc.),and calculates the degree of deviation from normal for the evaluation period by taking the average state of the same period of many years as the normal state.For the second category,the degree of drought was mainly measured by the anomaly degree of vegetation condition under drought stress,where the vegetation condition includes vegetation greenness indexes,vegetation moisture indexes,vegetation fluorescence indexes,and vegetation temperature indexes.The comprehensive monitoring method mainly measures the degree of vegetation drought by integrating drought-causing factors,vegetation conditions,and environmental parameters(e.g.,land cover type,agriculture irrigation,ecoregion,soil hydraulic parameters,etc.).After reviewing the current research status,we found that:(1)the overall development trend in the remote sensing of vegetation drought is'refinement','integration',and'intelligence',which means that the future remote sensing monitoring methods should be more faster,more sensitive,higher resolution,more comprehensive,smarter and so on;(2)The main challenges are:the remote sensing spatial resolution of vegetation drought background and characteristic parameters is still coarse,and the time series is still short and delayed.The existing vegetation drought remote sensing comprehensive monitoring models do not yet effectively integrate water,fluorescence,and temperature anomaly indicators.The response of vegetation canopy temperature,greenness,water,and fluorescence anomalies to drought stress has not been fully understood,and the establishment of existing drought remote sensing comprehensive monitoring models lacks the constraints or guidance of vegetation response laws to drought stress.(3)Future research directions include:Inversion and quality improvement of vegetation drought parameters with couple of mechanism model and learning model,vegetation canopy temperature anomaly detection technology based on two-stage trapezoidal model,and comprehensive monitoring model based on vegetation drought response law guiding deep learning algorithm.This study is helpful to break through the key technical bottleneck of drought remote sensing monitoring and accurately serve the needs of national disaster prevention and mitigation and ecological civilization construction.
NETL
NSTL
万方数据
