Analysis And Quantitative Estimation Of Relationship Between Land Cover And Soil Salinity In Coastal Areas Based On Near-Earth Spectral Imagery

The Yellow River Delta has a unique geographical location,rich natural resources and important ecological value.However,due to the influence of natural environment and human activities,the soil salinization in this region is serious,and the relationship between soil salinity and surface cover is obvious,which seriously restricts the development of regional agriculture.In this paper,the "Bohai granary" experimental area in Wudi County of Binzhou City was selected as the research area.The ADC multispectral camera and ordinary digital camera were used to obtain images.The two main parameters of soil salt content and vegetation coverage were selected to collect near ground multispectral photos,digital photos and synchronous measurement of soil salt data.The estimation model of soil salt was constructed and the distribution of soil salt was analyzed.It is of great value for the rapid and accurate monitoring of soil salinity and provides scientific guidance for the transformation of saline land.The main research contents and conclusions are as follows(1)The relationship between vegetation and soil salinity in different seasons was explored.Five vegetation indices were extracted from multispectral images,and there was a very significant relationship between each vegetation index and soil salinity.A total of 35 kinds of soil salt content estimation models were constructed.Through the model accuracy verification and optimization,the best estimation models of soil salt content in different seasons were obtained,which were Y=-0.597x+0.407 in spring,Y=-0.844x+0.651 in summer and Y=-0.242ln(x)+0.005 in autumn.The soil salt content estimation models of each season were practical.(2)The relationship between plants of different vegetation types and soil salinity was studied.The soil salinity estimation models of natural vegetation and planting vegetation were constructed by using vegetation index and measured value of soil salinity through multispectral images.After model optimization and verification,the best soil salinity estimation models were obtained,which were Y=0.057-0.299/x for natural vegetation and Y=-0.529x+0.920 for planting vegetation.(3)Based on digital images,the relationship between vegetation coverage and soil salinity was studied,and the best estimation model of soil salinity based on digital images was Y=-0.792ln(x)+1.199.The advantages and disadvantages of the two camera modes are compared and analyzed,which provides a new research idea for near earth remote sensing.(4)The distribution of soil salt content in the study area was analyzed by inverse distance weighted interpolation(IDW)of arcgis10.1.The soil salt content was divided into six grades:non salinization,very mild salinization,relatively mild salinization,moderate salinization,high salinization and saline soil.The results showed that the soil salt content in the northeast of the study area was the highest,and gradually decreased from northeast to southwest.In spring,the soil salt content in the Northeast was significantly higher than that in the southwest.In summer,the soil salt content decreased significantly from the southwest to the northeast,and increased significantly in autumn.