Informations et ressources scientifiques
sur le développement des zones arides et semi-arides

Accueil du site → Doctorat → Chine → 2020 → Study on the Remote Sensing-Based Measurement Model of Vegetation Resilience

China University of Mining and Technology (2020)

Study on the Remote Sensing-Based Measurement Model of Vegetation Resilience

徐雅晴;

Titre : Study on the Remote Sensing-Based Measurement Model of Vegetation Resilience

Auteur : China University of Mining and Technology

Grade : Doctoral Dissertation 2020

Université : China University of Mining and Technology

Résumé partiel
Vegetation can maintain the stability of the entire ecosystem.Understanding the connotation of the vegetation resilience and quantitatively measuring the resilience can provide a theoretical basis for ecosystem management and ecological restoration.However,the existing measurement systems of the vegetation resilience lack unified monitoring indicators and calculation methods.Meanwhile,some measurement methods don’t take the environmental variables into consideration and the results are subjective.Therefore,it is urgent to establish a vegetation resilience measurement system that can describe the dynamic changes of vegetation groups and the response characteristics of vegetation on disturbance.The multi-resolution,multi-temporal and multi-band of remote sensing images can meet the demands of land-surface monitoring.However,the number of observation platforms and data are increasing.Exploring multi-platform data fusion methods and establishing effective monitoring indicators are the keys to establish the vegetation resilience measurement system.Therefore,the general objective of this dissertation is to establish the monitoring indexes for vegetation ecosystem and remote sensing-based measurement model of vegetation resilience on the basis of the definition of vegetation resilience.In this dissertation,firstly,the vegetation disturbance factors and the influencing mechanism are summarized by the literature review.Then the mathematical modeling and statistical analysis methods are used to establish the remote sensing measurement model of vegetation resilience.The analysis method of resilience driving factors is also proposed.The drought monitoring index based on evapotranspiration has been proposed and been verified by using the climatical transect analysis method.For the noises and missing values in the time series data among model variables,the reconstruction methods are developed.Finally,taking the Northern Australia Tropical Transect,Shanxi Yicheng farmland ecosystem and Shaanxi Qinling forest ecosystem as the study areas,the model has applied to measure the vegetation resilience and explore the influencing factors of the vegetation resilience.The main conclusions are as follows :(1)The autoregressive model based on time series data can describe the dynamic changes and the memory effect of the vegetation ecosystem.According to the disturbance factors and the response of vegetation on water variation,this dissertation establishes the vegetation resilience measurement model based on ARx and lag correlation coefficient by taking vegetation index anomaly,temperature anomaly and drought monitoring index as model variables,and verifies the validity of the model through the variables correlation coefficient and stationarity test.Meanwhile,the Coupla function is used to establish the occurrence probability of extreme weather,and the driving factors of vegetation resilience changes are analyzed according to the tail correlation.(2)The drought index is an essential variable in the vegetation resilience model.Based on the land surface-atmospheric water balance process,a drought monitoring index CWDa based on the water deficit is proposed in this dissertation.The observations from the flux towers and the Australian Water Availability Project are used to calculate the CWDa.And the drought monitoring availability of the CWDa is tested in the Australian rainfall gradient zon

Mots clés : vegetation resilience; measurement model; remote sensing; drought monitoring index; reconstruction of time series data;

Présentation (CNKI)

Page publiée le 3 juin 2021