Font Size: a A A

Spatial Prediction Of Topsoil Texture And Analysis Of The Causes In Loess Region Of Southern Ningxia At Different Scales

Posted on:2021-02-14Degree:MasterType:Thesis
Country:ChinaCandidate:Z ShenFull Text:PDF
GTID:2393330602993229Subject:Agriculture
Abstract/Summary:PDF Full Text Request
Soil texture is one of the most basic physical properties of soil.It affects the air permeability,water retention and fertility of the soil,and determines many other soil physical and chemical properties.There are scale effects on the spatial variation of soil texture.At different spatial scales,the main influencing factors and the spatial variation pattern of soil texture are different.The Loess Pleteau in China is faced with a series of eco-environmental problems,such as severe soil erosion,decreases in soil fertility and productivity.Understanding the spatial distribution of soil texture and its influencing factors can help provide scientific basis for the precise management of farmland fertilization and irrigation,soil quality evaluation and prevention of soil erosion.However,at present,there are relatively few studies on the spatial distribution of soil texture in the loess area,and there is a lack of research on spatial prediction and cause analysis of soil texture at different scales.Therefore,in this paper,the loess area in southern Ningxia with complex terrain is selected as the research area to predict the spatial distribution of soil texture at the mesoscale of the county and the large scale of southern Ningxia.Empirical Bayesian Kriging with the symmetry logratio transform(SLR-EBK),regression Kriging with the symmetry logratio transform(SLR-RK)and random forest with the symmetry logratio transform(SLR-RF)were used to predict spatial distribution of soil particle size distribution in Haiyuan County based on the training set of 100 samples,combined with topographic factors,normalized difference vegetation index,soil types and land-use types.Then the predictions were validated with a validating set of 24 validation samples in the study area for comparison of prediction accuracy.By analyzing the influence of topographic factors and the differences in soil particle size distribution in different soil types,the causes of the spatial distribution of soil texture at the mesoscale are discussed.Classification and regression tree(CART)and random forest(RF)were used to predict spatial distribution of soil texture in southern Ningxia based on the training set of 343 soil profiles during the second national soil survey,combined with the same auxiliary variables.Then the predictions were validated with a validating set of 85 soil profiles in southern Ningxia and 122 samples in Haiyuan County for comparison of prediction accuracy.By analyzing the influence of topographic factors and the differences in soil texture in different soil types and land-use types,the causes of the spatial distribution of soil texture at the large scale are discussed.The conclusions are as follows:(1)Random forest(RF)is the optimal fitting model for prediction of distribution of soil particle size distribution in Haiyuan County and spatial distribution of soil texture types in southern Ningxia.(2)Results predicted by three methods show a spatial distribution trend that the sand content is lower in the southwest and higher in the northeast of Haiyuan county,while the silt content and clay content are higher in the southwest and lower in the northeast.(3)In the mesoscale loess area(Haiyuan County),elevation(Ele)is the most important auxiliary variable for predicting the spatial distribution of soil particle size distribution,followed by soil type(ST).The variation is mainly affected by natural conditions.The higher the altitude,the heavier the soil texture.Dark loessial soil,gray cinnamon soil and red clay are obviously heavier than other soil types.(4)The soil texture in southern Ningxia is mainly light loam,and it's heavy in the south area but light in the north area.(5)In the large-scale loess area(southern Ningxia),soil type(ST)is the most important predictor variable,followed by topographic factors,and finally land use type(LU),indicating that the spatial variability of soil texture in southern Ningxia is the result of a combination of factors.The soil texture of aeolian sandy soil and sierozem is lighter,while the soil texture of dark loessial soil and gray cinnamon soil is heavier,and the alluvial soil is mainly sandy loam and light loam.The higher elevation(Ele),the heavier the soil texture.The effect of anthropogenic cultivation on soil texture in southern Ningxia is mainly reflected in low-altitude areas.The soil texture of dry land is slightly stickier than that of grassland.The woodland is affected by natural conditions such as vegetation and moisture,and its texture is relatively heavy.Paddy fields and beaches are affected by flooding,and heavy loam accounts for a relatively large amount.Most of the unused soil is sandy.
Keywords/Search Tags:Loess area in southern Ningxia, Soil texture, Random forest, Spatial prediction, Analysis of causes
PDF Full Text Request
Related items