Study On Remote Sensing Inversion Of Texture Of Cultivated Soil And Countermeasures Of Soil Fertility Improvement In Youyi County

Arable land is an important resource to support social development,the basis for guaranteeing agricultural production,and the foundation for guaranteeing national food security and the quality and safety of agricultural products.Arable land strength,on the other hand,is the cornerstone for achieving high and stable yield of food crops,which is related to the sustainable and healthy development of society and the material life needs of the people.Soil texture is one of the important indexes to evaluate cultivated land fertility.Soil texture is a very stable natural and physical property of soil,which has an important influence on soil structure,porosity,fertility,water retention and tillage.The spatial distribution of soil texture in the region is of great importance to the evaluation of arable land quality grading,the dynamic change of arable land quality and the efficient use of arable land resources.Using remote sensing technology to invert soil texture can solve the problems of traditional methods relying on field work,strong subjectivity,time-consuming and laborious.Therefore,remote sensing technology is an effective method to obtain the spatial distribution of soil texture accurately and quickly.In order to analyze the productivity of arable land in black soil areas and provide accurate and quantitative data to support black soil protection,this study inverts and maps the results of soil spatial distribution in Youyi County.Firstly,to improve the accuracy of soil texture inversion,a total of 25 available Sentinel-2 images from 2019-2021 were selected to find the most suitable time window for soil texture inversion in the region and to analyze the reasons affecting the inversion accuracy.Secondly,suitable Sentinel-2 and Landsat 8 multispectral remote sensing images(total 6images)were selected within the optimal time windows of 2019,2020 and 2021,and a random forest model was used to compare and analyze the model accuracy of different inputs.At the same time,the temporal information is introduced to analyze the inversion accuracy of soil texture based on multi-temporal and multi-spectral data,and the inversion of spatial distribution of soil texture in Youyi County is realized.Finally,in order to analyze the relationship between soil texture and cultivated land fertility in Youyi County,the influence of soil texture on soil nutrients and the linkage between soil texture and crop growth in the region were analyzed,the key areas with low arable land strength grades in the study area were monitored,and countermeasures to improve the ground strength in the region were suggested.The main results of this study are as follows:(1)The best time window for remote sensing inversion of soil texture in Youyi County is from late April to mid-May,where the main reasons affecting the change of inversion accuracy are soil moisture and straw cover.This study found that the accuracy of the soil texture inversion model based on Sentinel-2 remote sensing images obtained within the time range of late April-mid May in the study area was high,and therefore,late April-mid May was considered as the time window for remote sensing inversion of soil texture in Youyi County.Meanwhile,soil water content and straw cover are important reasons for the accuracy of soil texture inversion.Soil moisture and straw cover affect the accuracy of remote sensing inversion of soil texture in the visible,near-infrared and short-wave infrared bands(VNIR/SWIR,0.4 to 2.5μm)region.And only images with relatively low LSWI value and NDTI value at the same time can produce better soil texture inversion accuracy.(2)Sentinel-2 and Landsat 8 multispectral images in the short-wave infrared 1 band and short-wave infrared 2 band are the sensitive bands for soil texture inversion.Soil texture is sensitive to the spectrum in the visible near-infrared band-shortwave infrared band,especially in the shortwave infrared band,which has a unique spectral reflectance response band,so it can use regional spectral characteristics to predict soil texture content.Among the six images from 2019 to 2021,band 12(band 7 of Landsat 8)and band 11(band 6 of Landsat 8)of Sentinel-2 image are sensitive bands for soil texture inversion in Youyi County,and the sensitive bands of the two sensors correspond to shortwave infrared 1 and shortwave infrared 2 bands,respectively.(3)Using spectral index combined with sensitive band as input can improve the inversion accuracy of soil texture compared with using only single band of image.The multi-spectral images such as Sentinel-2 and Landsat 8 used in this chapter have few bands.The accuracy of soil texture inversion is improved by constructing spectral indexes to varying degrees.The R~2 of silt inversion is increased by 0.001～0.190,and RMSE is reduced by 0.01～1.759.The R~2 of clay inversion increased by 0.016～0.115,and RMSE decreased by 0.048～0.331.The R~2 of sand inversion increased by 0.039～0.117,and RMSE decreased by 0.509～1.410.In the random forest model,the difference index constructed by the shortwave infrared 1 and shortwave infrared 2 bands with the blue,green and red bands is the most important input.Among them,the spectral index D7＿2(D12＿2)is an important spectral index in each random forest model for soil texture inversion.(4)The stability and accuracy of the single-period and multi-period models established using Sentinel-2 and Landsat 8 data can meet the requirements of rapid monitoring of soil texture.Due to the different accuracy of soil texture inversion in different years and different sensors,the stability and accuracy of the two-period image inversion model based on Sentinel-2 and Landsat 8 are higher than that of the single-period image in 2019-2021.From 2019 to 2021,the R~2 of silt inversion based on two images was 0.673-0.793,and RMSE was 7.624-7.929.The R~2 of clay inversion was 0.724-0.770,and RMSE was 1.569-1.720.The R~2 of sand inversion is 0.703～0.742,and RMSE is 7.964～8.537.Among them,the accuracy of inversion of clay content based on multidate images in 2019is the highest,and the accuracy of inversion of silt and sand based on multidate images in 2020 is the highest.(5)The spatial distribution of soil texture in Friendship County was obtained.Through remote sensing inversion of the spatial distribution of soil texture in Youyi County,it can be seen that the areas with high silt content in the study area are mainly:northeast,northern and southern small areas;the spatial distribution of clay content is consistent with that of silt content.The distribution of sand content is opposite to that of silt and clay.The sand content is higher in the central part of Youyi County(near the main urban area),the western mountainous area and the south,especially in the central part of Youyi County.There is a clear’sand circle’in the field area.(6)The spatial distribution of soil texture affects the grade of cultivated land fertility in Youyi County.It is mainly reflected in two aspects.On the one hand,On the one hand,soil texture,which is an important evaluation factor for arable land strength evaluation,affects soil nutrients.The contents of clay and silt grains were significantly and positively correlated with soil organic matter,p H,total nitrogen,alkaline soluble nitrogen and fast-acting potassium,and negatively correlated with total phosphorus and potassium;the opposite was true for sand grains.Soil texture,together with these factors,affects the level of arable land strength.On the other hand,crop growth,which characterizes the level of arable land strength,is influenced by the spatial distribution of soil texture in the study area.Good crop growth in the northeast,north and a small part of the south of Friendship County where the content of powder and sticky grains is high;In the central region with high sand content,the crop growth was poor,especially in the"sandy circle"area with high sand content in the classification results of 2021 crop growth.(7)Suggest countermeasures for key areas within Friendship County.By analyzing the inverse spatial distribution of soil texture in Youyi County,it was found that the central part of Youyi County has more sandy soils,which affects the ground strength level of cultivated land and causes unsatisfactory crop growth conditions in the area.Therefore,this study suggests countermeasures for applying improvers,optimizing fertilizer application techniques and adjusting cropping structure to improve arable land strength grade and crop growth capacity.