Font Size: a A A

Study On Inversion Of Soil Salinity In Dengko Area Hetao Basin Based On RS

Posted on:2021-03-21Degree:MasterType:Thesis
Country:ChinaCandidate:Z K YangFull Text:PDF
GTID:2370330602972236Subject:Geological Engineering
Abstract/Summary:PDF Full Text Request
The ecological protection and high-quality development of the Yellow River Basin has risen to the national strategy in 2019.Agriculture is one of the Hetao Basin primary economic source,soil salinization is one of the major issues restricting the development of local agriculture.Through remote sensing technology,we can determine the salinity of the soil,add sedimentary phase factors and land use factors to analyze its impact on the distribution of soil salinity.The research can provide basic data and sedimentary facies for the soil management of relevant departments.Use remote sensing to interpret the land use status and analyze the distribution law of soil salinity.The main knowledge and results achieved in this article are as follows:(1)Optimize the model and invert the soil salinity in the study area.After comparing the data,we can know that the BP neural network has a better fitting effect in the study area.Using Sentinel-1A radar data after preprocessing such as multi-view,filtering,correction,etc.,a 2-7-1 BP neural network model was established.After comparing with the sample points,the root mean square error RMSE = 0.7538,R2 = 0.7357,we can use this BP neural network model for salt inversion in research area.(2)Use preliminary classification and special classification system to carry out remote sensing geological interpretation and land use status interpretation in Hetao Dengkou area.In the study area,a random forest machine learning algorithm is used to supervise and classify various types of sediments and surface cover units,combined with the spectral characteristics of specific sediments to carry out special interpretations of alluvial fans,ancient river channels,etc.,then get the latest geological interpretation and the current status of land use.The results show that the surface of the study area is mainly sediments from the Dishaogouwan Formation of the Upper Holocene,with artificial transformation and severe salinization.(3)Comprehensive analysis of the spatial distribution of soil salinity combined with sedimentary facies and land use status.The soil salinity of the alluvial plain is concentrated in the crescentic lake subfacies,natural levee microfacies and river floodplain facies.In the mountainous areas where the height difference is obvious,the soil salinity is concentrated in valley ravines and low-lying areas.(4)According to the study of salt distribution law and influencing factors,this paper proposes measures to prevent soil salinization.Scientific irrigation methods can effectively reduce the degree of salinization,and flooding irrigation will cause secondary salinization in farmlands around villages and towns.Through inversion and comprehensive research on soil salinity in Dengkou area,the influencing factors and spatial differences of soil salinity are analyzed,which can provide reference for relevant departments to carry out land management and planning.It has certain practical significance.
Keywords/Search Tags:Remote sensing, soil salinity inversion, radar image, neural network model, Hetao Dengkou area
PDF Full Text Request
Related items