Font Size: a A A

Application Of SWAT Model To Study The Water Budget Of Forest And Grass Ecosystems In Hilly Region Of Red Soil In Northeast Jiangxi Province

Posted on:2005-10-08Degree:MasterType:Thesis
Country:ChinaCandidate:J F DaiFull Text:PDF
GTID:2133360125469198Subject:Soil science
Abstract/Summary:PDF Full Text Request
Because of the concentrated rainfall and irrational land use, the seasonal drought, soil and water losses also exist in hilly region of red soil in China. It is important to study the water budget in hilly region of red soil. By combining the experiment, field observation and data collection with SWAT model simulation, the water budget in different types of land use (forest and grass ecosystems, orchard and crop ecosystems) were calculated by SWAT model. Then the water circulation and water budget in the following years were predicted and the main factors that influenced the water budget in hilly region of red soil were analysised.The characteristics of the temporal and spatial distribution of the surface runoff, percolation from the root zone, evapotranspiration and soil water storage of the natural grassland, board-leaved forest land, mixed forest land and coniferous forest land of four catchments in the red soil hilly region were studied for the first time using SWAT model in China. The outputs of SWAT model were calibrated and valided against observed surface runoff. Comparing simulated monthly surface runoff to observed one, a good result was gained with a R2 of 0.90 and an Ens of 0.74. The case study illuminated that SWAT model can be used to simulate the surface runoff, percolation and evapotranspiration of different forest and grass ecosystems in red soil area.The results showed the variations of the rainfall within a year also lead to the difference of the runoff, percolation and soil water storage among the months in the year. The surface runoff and percolation often occurred in rainy season. The rates of three months (Apr. to Jun.) surface runoff and percolation in annual ones were all above 50%. The comparison among the four ecosystems revealed that board-leaved forest land and mixed forest land concentrated highter percents of runoff in rainy season than the coniferous forest land and natural grassland. The rate of four months (May to Aug.) evapotranspiration in annual ones was above 55%. The annual fluctuation of the weather conditions, especially the rainfall, resulted in the fluctutation of the water budget. The runoff and percolation were effected mostly by the rainfall, whereas the evapotranspiration and the soil water storage were less effected.There were differences in the water budget of four forest and grass ecosystems. Comparing with the natural grassland, the forest land reduced the runoff more effectively, in which the board-leaved forest land and mixed forest land were most preferable. Theevapotranspiration was the largest component of the water budget in the forest land. And the evapotranspiration of the forest land was larger than natural grassland.SWAT model reflected the effect of the landuse changes to the regional water budget. The results indicated that the runoff of the orchard or crop ecosystem was bigger than that of the forest land. The model predicted the influence of weather conditions in the future on water budget based on the past weather datas. The results showed that weather conditions were the most pivotal factor that influenced the long-term water budget in the region scale, while the land use was also important to the water budget. Forest land would change the regional water budget more remarkably with the time, however the influence of the natural grass on water budget was smaller.The sensitivity of parameters on water budget in different forest and grass ecosystems were studied using single-dimensional sensitivity analysis. The result showed that parameters' sensitivity was different in forest and grass land. In forest land the sensitivity factor of leaf area index and soil layer depth was bigger than soil physical properties such as the permanent wilting point, saturated conductivity and available water capacity, but reversed in grass land. The maximum rooting depth and the soil albedo were most sensitive to water budget in grass land, but not in forest land.
Keywords/Search Tags:SWAT model, water budget, forest and grass ecosystems, red soil
PDF Full Text Request
Related items