| Using 3S technology with integration multiple factors of soil formation, to quantitative analysit soil ecological environment objectivly. And based on the conclusion, the comprehensive cartographic of soil nutrient spatial distribution characteristics have been finished, and it can maximum close to the natural state, and it will povide an important foundation to sustainable utilization of soil resources, an objective evaluation of the soil fertility, fertilizer zoning and promotion of measuring technology of formula fertilizationIn this study, we make the farming area in Lantian County, Shaanxi Province as the research area.A set of mapping implementation for regional soil nutrient spatial distribution plan which is more complete and scientific have been designed by using the theory and methods of 3S(GIS,GPS,RS) technology. Firstly, the implementation scheme of the rule sampling in study area have been designed by spatial analysis of GIS and spatial positioning function.Next, soil forming factors which effects of soil nutrient formation, including soil factor, topographic factors, biological factors, climatic factors have been collected as much as possible by using RS digital image processing and GIS spatial analysis, and two types methods of GIS spatial analysis have been designed to quantificat this factors. Thirdly, to complete single soil nutrient multiple regression prediction mapping scheme, soil nutrient spatial regression prediction have been carried through multiple linear regression analysis method with integrating all the strongly related influence factors, and comparison the accuracy of conclusion with the traditional Kriging interpolation method to test effective of this method.Finally, In order to designing the regional soil nutrient comprehensive mapping program, the soil nutrient comprehensive mapping in the study area have been made by using the principal component analysis method based on the rusult data of the individual soil nutrient which get through regression forecasting.The result shows as follows:(1) Soil factor collection and quantification. The soil types, human activity index, vegetation coverage, drought degree index, elevation factor, slope factor, slope aspect factor and Topographic Wetness Index were collected by the principle and method of RS digital image processing and GIS spatial analysis, based on the soil nutrient forming factors such as soil parent material, topography, climatic factors, biological factors. Two methods for the quantification of soil factors were designed based on the theory and method of GIS and RS. Method one, some data range would be stretched to the range between 0 and 1,which data can charact the surface and belong to relative measure index such as vegetation coverage factor, drought index and Topographic Wetness Index. Method two, some polygon data such as soil type, human activity index and some topographic distribution data like elevition factor, slope factor and slope aspect factor would be update by the weight of soil nutrients forming, which weight could be get by local statisticing the average nutrient value of sampling point in polygon.The soil factors are unified by using these two methods, and the two different result are analyzed and evaluated in the process of modeling.The results showed that the second method is better than method one, due to the result of method two full express the influence degree which effected by the soil forming factors on the formation of soil nutrients.(2)The spatial distribution characteristics of soil nutrients were analyzed by building the regression prediction model by AcrGIS under the influence of multiple factors. The result showed that the forecast accuracy between the method of multiple regression prediction and the traditional method of Kriging interpolation prediction were quite similar. But the advantages of multiple regression prediction method is difficult to achieve in traditional Kriging method. There are some advantages in the results of linear regression prediction model. Fistly, this method could have integrate all soil formetion factors which have significant correlation to soil nutrient into the soil nutrient spatial characteristics prediction.Secondly, the linear correlation between the impact of soil forming factor and soil nutrient content was more significant, and the accuracy of regression prediction mapping was higher; The lastly, in the result of regression prediction, the change regulation of soil nutrients spatial distribution characteristics is similar to the change regulation of soil forming factors like soil type, land use, topography and the combination of water and heat conditions, and it overcome the bug in the result of traditional Kriging interpolation method which have some phenomenon like the distribution patch shape.The results of regression prediction include the spatial information of various kinds of soil forming factors.The spatial distribution of soil nutrients is the most close to the nature situation from the visual sense, which makes the result more reliable.(3) The comprehensive mapping of soil nutrient have been made by the model of principal component analysis which build by GIS. The result showed that 4 soil nutrients spatial distribution characteristics derived from regression prediction have been stretched to between 0 and 1 by pixel linear stretching method. Every soil factors could have been impartial by this scheme in the program of PCA analyst, and the result of comprehensive mapping is seem to objective and scientific. The result derived from this scheme have overcame the phenomenon that composition information unequal extraction due to different of the unit of soil nutrition before factors stretched. Owing to the datasource of comprehensive mapping is the result of spatial regression prediction, the spatial distribution characteristics of soil nutrition comprehensive mapping is same to the spatial characteristic of soil type, human activity intensity, topographic change and water and heat combination condition.(4) High value of soil nutrient content in the study area mainly distributes in this region such as the soil type is alluvial soil, paddy soil, lou soil, standing stubble soil, flat terrain, the slope is between 0 and 8 degree, the slope is mostly east slope and southeast slope, the condition of water and heat combination is good. The specific region is distribution in the town of Languan, Jinshan, San Guanmiao, Hou, Huaxu, Sanli, Xiehu, and this region is concentration on a zonal distribution in agricultural area along the river. This area is flat, gentle slope, and the soil nutrient concentrated along the river was migrated down by water from from higher terrain, steep regional.The main types of agricultural planting in this area is kiwi fruit, grape and rice irrigation fertilizer and so on.The soil nutrient content of this area is relatively higher than the other regions, due to this area have some superior soil conditions like superior combination of water and heat, and frequent, strong human agricultural activities. The low value areas of soil nutrient content mainly distribut in this region such as marsh soil, purple soil, higher terrain, steep slope, slope aspect to the southwest, west, and the combination of water and heat conditions in poor. This area specific distribut in the town such as Xiaozhai, Jiu Jianfang, Huaxu, the central of Xiehu, and central of Sanli, which near the mountain in southeastern of Lantian County. And this area of low value soil nutrient mainly distribute in area near Qinling Mountains, northwest and highly topography, steep slope at the central of study area, there are serious soil erosion and difficult soil nutrient enrichment. The development of forestry animal husbandry, grass, soil and water conservation, comprehensive utilization was as Agricultural planning and management measures in previous years. The causes of the low soil content is that this area have poor natural conditions, like serious erosion of water, poor of combination of water and heat conditions, and the weak of human agricultural activities. The process of soil nutrien formation have been changed by human activies, and it lead to the result of the high value area of the soil nutrient content mainly distributed in arable land. The soil nutrient in other land types such as grass and forest is relatively to higher than sand and water. The results can provide theoretical and technical reference for soil nutrient comprehensive digital mapping. |
PDF Full Text Request