Lmpacts Of Climate And Vegetation On Surface Water Balance In Poyang Lake Basin,China

Recently, the hydrological cycle has changed significiantly, with climate change and increase of CO2concentration. However, as an interactive part of the earth-atmospheric system, the influences of vegetation physiology and structure (such as stomatal and LAI) on the hydrological cycle have been paid less attention or been neglected, and made the calculations of the hydrological components (such as streamflow and evaportranspiration) and the causes of the hydrological cycle changes to be inaccurate. Because of no considering the dynamic time-series of CO2concentration and the differences in responses of vegetation (stomatal and LAI) to CO2, SWAT (Soil&Water Assessment Tool) model was improved preliminary on the module of vegetation growth. Taking four typical subbasins of Poyang Lake Basin as the study areas, the simulation capablities were evaluated through comparing the hydrological processes, such as streamflow and evaportranspiration. By contriving several experiments, the contributions of climate variables and vegetation factors to changes of streamflow and evaportranspiration (1971-2005/2003) were calculated quantitatively with the improved model. In the last, using IPCC-4different models'three scenciors (SRESA1, SRESA2and SRESB1) datasets in future (2011-2090), the changes of streamflow and evaportranspiration relative to the reference period (1971-1999) were predicted. Results were as follows:(1) The variation characteristics of hyro-climate variables from1961to2005(2003) were revealed. During1961to2005(2003), climate variables and hydrological components (including streamflow and evaportranspiration) existed obvious changes in Meigang, Saitang, Gaosha and Xiashan, respectively. In all the basins, annual precipitation and streamflow showed increasing trends with decreasing evaportranspirarion at different ratios. Thus, the changes of other climate variables were basically in accordance. For mean and minimum temperature and relative humudity, they all had consistent and significiant increase. The maximum temperature in Meigang, Saitang and Xiashan basin expressed slight increase, but there was slight decrease in Gaosha. Wind speed at2m and net solar radiation in each basin showed consistent and significiant decreasing trends.(2) SWAT model was improved and evaluated in simulation capabilities of the hydrological processes. Original and improved SWAT model were calibrated based on the auto-calibration function. The monthly and the yearly streamflow (actual evaportranspiration) correlation coefficients (R) between observations and simulations from the original model and the improved one were all more than0.90. Using the Nash-Suttcliffe efficiency coefficients (Ens) as evaluation criteria, the parameters from the original model and the improved model in simulating streamflow of four basins were basically in accordance at level of Class A or Class B. Comparing the mean relative errors(MRE) of monthly and yearly actual evaportranspiration, MRE values form the improved model were smaller than ones from the original model. It was essential for evaluating the actual evaportranspiration exactly to calculate the water resources. Taking the streamflow-assurance as premise, the accuracy of actual evaportranspiration simulations with the improved model was improved significantly. Then, the improved model could support a tool for evaluate the contributions of climate and vegetation changes to streamflow quantitative.(3) Based on the improved SWAT model, the contributions of climate and vegetation to changes of streamflow and evaportranspiration were calculated quantitatively by contriving several experiments. In order to separate the contributions of every influence factors to annual streamflow (annual actual evaportranspiration) changes, two methods (Case A and Case B) were given and compared. Case B was more reasonable to calculate the contributions of every influence factors than Case A. From1971to2005(2003), the main reason of decreasing actual evaportranspiration was the reduction of wind speed with more than69%of the trends in each basin, especially accounting for133.03%and119.22%in Meigang and Gaosha, respectively. The vegetation changes (including stomatal conductance and vegetation LAI) had some contributions to the decrease of the actual evaportranspiration, but the values were obviously smaller than other factors'. Simultaneously, the increases of streamflow for all three basins (Meigang, Saitang and Gaosha) were mainly because of increasing precipitation, with the percent contribution of more than70%, especially accounting for88.79%in Meigang Basin. However, actual evaportranspiration change was the main reason of streamflow trend. For each basin, the role of vegetation changes (including stomatal conductance and LAI) was quite limited with the contribution less than0.05m3s-1year-1.(4) Taking the three scenariors (SRESA1B, SRESA2and SRESB1) of climate during2011to2090downscaled based on Delta method and WXGEN weather generator as input field, the changes of hydrological cycle relative to the reference period (1971-1999) were predicted. Comparing the multi-model climate variables, actual evaportranspiration and streamflow during2011to2090with the mean from1971to1999, they all will change obviously in some extent. Generally, the multi-model averaged annual precipitation, actual evaportranspiration and streamflow will decrease, increase and decrease respectively, but the change ranges are different among the three scenarios, of which the ranges of SRESA2will be significiant.