Desert Environment Monitoring Using Multi-source Remote Sensing Images Based On Remote Sensing Ecological Index

With the development of social economy and the acceleration of urbanization,the ecological environment is facing great pressure.The increase of human activities,the destruction of ecological environment and the large-scale desertification of land causing frequent dust weather and even dust storms have seriously affected people's life.In order to prevent and control land desertification,a series of ecological restoration policies were introduced to control ecological degradation.Ecological environment monitoring using remote sensing technology can dynamically monitor the environmental change in the ecological restoration project,but there is a lack of quantitative evaluation on the degree of ecological restoration and environmental desertification.In this study,the Mu Us Desert was selected as the study area,and Landsat satellite images and MODIS multi-source remote sensing data were used to analyze the ecological and environmental conditions of Mu Us Desert from 2000 to 2020,so as to obtain comprehensive indicators that could represent the environmental conditions for ecological assessment.The main contents and conclusions are as follows:1.The remote sensing ecological index(RSEI)based on the Pressure-State-Response(PSR)framework can reflect the regional ecological environment well but hard in continuous monitoring.In this study,the dryness index and principal component analysis were improved,the soil index was used as the dryness index to establish the model with high accuracy,and the RSEI index was calculated by unifying the direction of eigenvector in the principal component analysis.The improved RSEI model is scientific and accurate.2.From 2000 to 2020,the four ecological indexes,vegetation index and land surface humidity showed an upward trend,while the soil index showed a downward trend.The change of land surface temperature fluctuated greatly and the trend was not obvious.Four indexes are normalized to unify their dimension.The trend of the four ecological indexes after normalization is different from that before normalization.It is considered that the overall distribution of the index has little change and the rise of the average value of the index is the result of the rise of some pixel indexes.Among the four ecological indexes,moisture index and soil index have a high negative correlation.The moisture and dryness of the ecosystem represented by them are opposite,but their specific meaning and ecological information described by them are different.3.The applicability of the vegetation index,humidity index and surface temperature is demonstrated by using the data of MODIS products and meteorological station.It is concluded that the NDVI has a high consistency with the results of MODIS products.The land surface humidity and precipitation show a rising trend in the past 20 years,and there is a high correlation between land surface temperature,MODIS temperature products and average air temperature.4.The principal component eigenvectors are in opposite directions,and the results are comparable after unification.The remote sensing ecological index RSEI calculated by the principal component analysis showed an overall upward trend,and the factor synthesis score obtained by the factor analysis also showed an overall upward trend in the past 20 years,indicated the improvement of the ecological environment.The changes of RSEI index from 2010 to 2015 and from 2017 to 2020 are consistent with the changes of synthesis scores,all go from steady to decline,indicating that the ecological and environmental conditions were deteriorated from a relatively stable state during the decade from 2010 to 2020.5.Principal component analysis(PCA)and factor analysis(FA)have different emphasis.According to the variation results of RSEI index,the contribution of land surface temperature to PCA is greater,while the variation results of factor synthesis score indicate that correlation and commonality of data contribute more to factor analysis.In conclusion,the improved RSEI index is suitable for application in the Mu Us Desert area.The ecological environment in the study area has been improved from 2000 to 2020 to some extent,the implementation of ecological restoration policies and artificial desert control have played a considerable role.However,the analysis shows that after 2016,the ecological environment has a trend of degradation,and it is still necessary to continue to strengthen the ecological restoration of the desert.