Identification Of Vegetation Stress Type Based On Spectral Characteristics

Drought and heavy metal pollution are two typical stresses.Plants of one area may suffer from many kinds of stresses.It has important significance to monitor vegetation stress accurately for vegetation management and ecological protection.Some researches focused on the plant response of biophysical and biochemical parameters,spectral parameters and correlation analysis under single stress type in different concentrations,but there was less reseach on how to distinguish stress type in the area with kinds of stress factors.Hyperspectral data has good application in information extraction of vegetation stress because of its high spectral resolution.Bases on maize pot experiments,this paper contrasted the difference of plant parameters and spectral parameters between CK,drought,copper and combined stress groups.Firstly,key plant parameters and key spectral indices were determined accorfing to the changes of physiological parameters and their corresponding spectral index under different stress types.Secondly,the model of stress type identification was build under the premise of experiment condition.Lastly,satellite remote sensing image was used to verificate the condition of stress type identification.Results showed that:1)Chlorophyll content,leaf area,moisture content were selected as the three key bio-physicochemical parameters to identificate stress type through the data analysis of spectral data and physiological parameters.2)mND705,TCARI,SWISI were selected as the key spectral indices by relevance analysis and stress type identification model was constructed based on spectral indices.3)Hyperion images were processed and analyzed to identify copper stress and drought stress in Dexing copper mining area and the western suburbs of Changchun respectively,which verified the feasibility of identification of stress type based on spectral characteristics.