The Applications Of Remote Sensing Models Of Soil Salinization Based On Feature Space

Increased soil salinization poses a serious obstacle to achieving sustainable use of land resources.For different geographical locations and different climatic regions,a variety of two-dimensional feature space models are used to quantitatively assess the soil salinity conditions in typical salinized areas and to select the optimal inversion model,which is of great significance for accurate monitoring of soil salinity content in the region.The applicability of the optimal monitoring model has also not been systematically compared in the more salinized arid and coastal regions of China.In view of this,the present study was conducted to address the land degradation problems arising from the Yanqi Basin(arid zone)and the Yellow River Delta(humid zone),which are typical salinized inland arid areas in China,based on field sampling data and Landsat 8 OLI image data of the same period.The Albedo,Salinity Index(SI),Modified Soil-adjusted Vegetation Index(MSAVI)and Normalized Difference Vegetation Index(NDVI)were selected as biophysical parameters of salinization processes.Three feature space salinity estimation models,Albedo-MSAVI,SI-Albedo and SINDVI,were constructed based on the correlation between the parameters,and the best model was selected for the inversion.The soil salinity class distribution maps for 2020 in Yanqi Basin in the arid region and for2019 in Kenli County in the Yellow River Delta in the coastal region were further obtained to analyses the salinity distribution characteristics and driving factors and to provide constructive suggestions for later improvement and management.At the same time,a comprehensive comparison of the optimal models in the two regions was carried out to realize the applicability of the optimal inversion model to typical salinization areas in different climatic zones(arid and humid zones)in China.The results of the study show that.(1)In this study,feature spaces were constructed using feature parameters that can better reflect soil salinity information,which can effectively distinguish between different levels of saline land in the study area and improve the accuracy of salinity information monitoring.By comparing the Albedo-MSAVI,SIAlbedo and SI-NDVI feature spaces,it was found that the SI-Albedo feature space performed best in monitoring salinity in the Yanqi Basin in the arid region and the Yellow River Delta in the coastal region,with inversion accuracy of 93.3% and 87.2%,respectively.It shows that the inversion salinity sensitivity indices are the same in both regions,and the eigenspace model constructed from the SI and Albedo can better reflect the land degradation characteristics of soil salinity in the study area,and has the potential to estimate soil salinity,and can be used as a universal salinization monitoring model in coastal and arid areas of China.It also shows that the feature spaces model approach has some advantages for the inversion of soil salinity in different typical salinized areas(coastal and arid areas).(2)The SI-Albedo feature spaces model with the highest inversion accuracy was used to invert the distribution of soil salinity in the two regions,obtaining the distribution of soil salinity levels in the Yanqi Basin in the arid region in 2020 and in Kenli County in the Yellow River Delta in the coastal region in 2019,respectively.The inversion results show that about 90% of the land area in Yanqi Basin has salinization.The spatial distribution of soil salinity was characterized by a gradual decrease in soil water content from west to east and from north to south,and a gradual increase in soil salinity.The degree of salinization in Kenli County was also heavy.About 70% of the land area was salinized.The distribution of soil salinity was characterized by a "high-low-high" distribution,with the eastern coastal areas,inland marginal areas and central areas in descending order.(3)To analyses the causes of soil salinity in the Yanqi Basin in the arid region and in Kenli County in the coastal region,seven driving factors(intensity of human activity,land use,elevation,depth of buried groundwater,mineralization,average annual precipitation and slope)were selected,taking into account the effects of climate type,human activity,topography and land use type.At the same time,in view of the different degrees of soil salinization in the two regions,the corresponding management measures are proposed from three levels: prevention of secondary salinization,improvement of salinization,and removal or transformation of salinization.Combining the hydraulic improvement measures,agricultural improvement measures,biological improvement measures and chemical improvement measures,the corresponding sustainable development countermeasures for the regions with different degrees of salinity are proposed,providing useful reference and reference for the management of soil salinization in the arid and coastal regions of China in the future.