Informations et ressources scientifiques
sur le développement des zones arides et semi-arides

Accueil du site → Doctorat → Chine → 2009 → Soil Nutrients Distribution and Simulation Withwepp Model in the Watershed of Loess Plateau

University of Chinese Academy of Sciences (2009)

Soil Nutrients Distribution and Simulation Withwepp Model in the Watershed of Loess Plateau

韩凤朋;

Titre : Soil Nutrients Distribution and Simulation Withwepp Model in the Watershed of Loess Plateau

Auteur : 韩凤朋;

Grade : Doctoral Dissertation 2009

Université : University of Chinese Academy of Sciences

Résumé
As one of the severe soil loss areas, it is very important to predict the soil and nutrients loss in Loess Plateau in future. The WEPP model can simulate and produce the spatial distribution of erosion sediment yield, which can be used for the land use design with the aim of reducing the soil and nutrients loss. So, it is vital to discuss the applicability of the WEPP model in the Loess Plateau.The study area is in the Liudao Gou watershed located in Shenmu County, Shannxi Province, China. The data needed for the WEPP model, the soil nutrients distribution and soil nutrients loss were analyzed by the field experiments and the laboratory analysis. By which, the applicability of the WEPP model in the different scales was evaluated. The conclusions could be drawn as follows :(1) The spatial variability of nutrients at the slope scale under different plant coverage and soil types. The purple medic can change the content of soil organic matter, nitrate nitrogen, and total nitrogen when the coverage is greater than 40 %. When the depth is deeper than 20cm, this effect was decreased. There are little differences in total phosphorus and ammonium nitrogen for different soil depths. The Stipa bungeana root was mainly distributed in the depth of 0-50 cm, and the root content decreased with soil depth. The trend can be simulated well by exponential model. The relationship between the soil nutrients content during the 0-50 cm and the root distribution can be simulated by linear model. The soil covered by Stipa bungeana can change the soil organic matter, total nitrogen, nitrate nitrogen content, especially in the depth of 0-20 cm. But this was not the case for the total phosphorus and ammonium nitrogen except for the surface soil layer. The content about soil total nitrogen, organic matter, total phosphorus, nitrate nitrogen, and ammonium nitrogen varied with different land uses. The total phosphorus, total nitrogen, and soil organic matter content in the forest, shrub land, and gully were generally higher than those in the grass land and farmland.(2) The nutrients content and spatial variability in the small watershed scale. All of the nutrients, especially for the total nitrogen and the ammonium nitrogen in the surface soil layer presented strong variability. The semivariances of organic matter, total phosphorus, and ammonium nitrogen can be simulated by Gaussian model, while total nitrogen and nitrate nitrogen can be simulated by exponential model. Most of the organic matter, total nitrogen and nitrate nitrogen content for the first meter in the watershed were distributed in the grassland, shrub land, and gully. The organic matter, total nitrogen, and proportion of nitrate nitrogen content for the farm land was greater than the proportion of farm land area. The total phosphorus and ammonium nitrogen were different from other nutrients, which were mainly distributed in the grassland, shrub land and gully.(3) The WEPP model can be used in the slope of the Loess Plateau. The WEPP model can be well used to simulate the annual precipitation. The month averaged rainfall from 2003 to 2008 was almost the same to the simulated rainfall except for April and August which were a little overestimated. For the manmade runoff plots, the simulated runoff and sediment yield was similar to the actual data. While for the natural plots, the simulated runoff was similar to the actual data but the simulated sediment yield was higher than the actual data. The Nash-Sutcliffe coefficient test showed that all of the simulated data were better than the average, with no significant difference by the T test. The simulated data about the nutrients loss in the manmade runoff plots was similar to the actual data ; while in the natural plots the simulated results was not accurately enough. However, the Nash-Sutcliffe coefficient test showed that the simulated efficiency in all of the plots were higher than zero. From these results we can know that the WEPP model can be used to simulate the water, soil, and nutrients loss at the slope scale.(4) The using of WEPP model in the watershed scale need deliberate. The sediments distributed mainly in the areas with higher altitude and sparse vegetations. From the intensity of soil loss, the WEPP model can simulate the soil loss intensity accurately. In this watershed the runoff in one year was 83m3, the sediment yield was 8.057t。But the simulated soil loss yield was higher than the actual data. Through the field experiment and laboratory analyses, the loss of organic matter, total nitrogen, total phosphorus, nitrate nitrogen, and ammonium nitrogen was observed to be 38.432kg, 4.014kg, 3.205kg, 0.102kg, 0.716kg, respectively. Since the WEPP model simulates the nutrients loss mainly based on the sediment yield, the simulated result was higher than the actual data.

Mots clés : WEPP model; Loess Plateau; soil nutrient; spatial variability;

Présentation (CNKI)

Page publiée le 16 mai 2018