Prediction Model of Soil Nutrients Loss Based on Artificial Neural Network

On the basis of Artificial Neural Network theory, a back propagation neural network with one middle layer is building in this paper, and its algorithms is also given, Using this BP network model, study the case of Malian-River basin. The results by calculating show that the solution based on BP algorithms are consis- tent with those based multiple - variables linear regression model. They also indicate that BP model in this paper is reasonable and BP algorithms are feasible.

Study on Relative Soil and Water Conservation Benefits of Ridge Tillage in Different Terrain Conditions in the Black Soil Area of Northeast China

Ridge tillage, which is a very common and important tillage measure in the black soil area of northeast China, has some soil and water conservation bene- fits, but has little attention. It is very important to explore the spatial distribution of the ridge direction of the arable land and its soil and water conservation benefits in different terrain conditions in the black soil area. So Binxian County of Heilongjiang Province was selected as the study area, and 168 field investigation units were ex- tracted by stratified sampling method and investigated. According to equations of slope gradient factor and slope gradient in ridge direction, and based on the soft- ware of Arcmap, SPSS and Excel, the investigation data of soil and water loss in Binxian County were analyzed and counted, The results show that in plain, hilly and mountainous areas, the average ground slope gradients are 1.92°, 6.20° and 8.27° respectively, and the average slope gradients along ridge direction are 1.33°, 4.52°and 6.45° respectively, which account for about 70%, 73% and 78% of the average ground slope gradients in the same terrain condition; the relative quantities of soil erosion in the present ridge tillage condition account for about 55%, 69% and 67% respectively of that in down-slope ridge tillage conditions, so the present ridge tillage has obviously relative soil and water conservation benefits. Based on these results, the reasons of the present ridge tillage status were analyzed, and some reform measures were proposed. The results could not only help to comprehend the spatial distribution and soil and water conservation benefits of ridge tillage in the black soil area of Northeast China, but also provide scientific references for the layout of local soil and water conservation measures.

Capacity of soil loss control in the Loess Plateau based on soil erosion control degree

The capacity of soil and water conservation measures, defined as the maximum quantity of suitable soil and water conservation measures contained in a region, were determined for the Loess Plateau based on zones suitable for establishing terraced fields, forestland and grassland with the support of geographic information system(GIS) software. The minimum possible soil erosion modulus and actual soil erosion modulus in 2010 were calculated using the revised universal soil loss equation(RUSLE), and the ratio of the minimum possible soil erosion modulus under the capacity of soil and water conservation measures to the actual soil erosion modulus was defined as the soil erosion control degree. The control potential of soil erosion and water loss in the Loess Plateau was studied using this concept. Results showed that the actual soil erosion modulus was 3355 t·km^(–2)·a^(–1), the minimum possible soil erosion modulus was 1921 t·km^(–2)·a^(–1), and the soil erosion control degree was 0.57(medium level) in the Loess Plateau in 2010. In terms of zoning, the control degree was relatively high in the river valley-plain area, soil-rocky mountainous area, and windy-sandy area, but relatively low in the soil-rocky hilly-forested area, hilly-gully area and plateau-gully area. The rate of erosion areas with a soil erosion modulus of less than 1000 t·km^(–2)·a^(–1) increased from 50.48% to 57.71%, forest and grass coverage rose from 56.74% to 69.15%, rate of terraced fields increased from 4.36% to 19.03%, and per capita grain available rose from 418 kg·a^(–1) to 459 kg·a^(–1) under the capacity of soil and water conservation measures compared with actual conditions. These research results are of some guiding significance for soil and water loss control in the Loess Plateau.

Cogitation on developing a dynamic model of soil wind erosion

Studies on soil wind erosion began with single factors affecting soil wind erosion; with increasing quantities of data being accumulated,the wind erosion equation(WEQ),the revised wind erosion equation(RWEQ),the wind erosion prediction system(WEPS),and other soil wind erosion models have been successively established,and great advances have been achieved.Here we briefly review the soil wind erosion research course and analyze the advantages and disadvantages of the current soil wind erosion models.From the perspective of the dynamics of wind erosion,we classified the factors affecting soil wind erosion into three categories,namely,wind erosivity factors(WEF),soil antierodibility factors(SAF),and roughness interference factors(RIF).We proposed the concept of a standard plot of soil wind erosion to solve the problem of uncertainty of the soil wind erosion modulus on a spatial scale,and provided methods to set similarity conditions in wind tunnel simulation experiments and to convert the spatial scale of the wind erosion modulus from the standard plot to a large scale field.We also proposed a conceptual model on the basis of the dynamics of soil wind erosion with the theoretical basis that wind produces a shear force on the soil surface.This shear force is partitioned by barely erodible soil surfaces and roughness elements on the ground,and the amount of soil loss by wind should be calculated by comparing the shear force of the wind on barely erodible soil surfaces with the anti-erosion force of the surface soil.One advantage of this conceptual model is that the calculated soil wind erosion modulus is not subject to changes of spatial scale.Finally,we recommended continual improvement of the existing models while also establishing new models.