| Water resources management strategy based on water demanding and Hydraulic engineering based on hydrological mechanisms have not coped with the increasing water resources demanding and associated environmental issues. Current research needs ecohydrology which consider the influence, interaction and evolution of ecological and hydrological processes as the theoretical basis of the integrated basin management and sustainable development.The upper reaches of the Heihe Basin is not only the regions of water resources generation and conservation, but also the most critical zones where the interactions and evolutions of ecological and hydrological processes occurs. The forest-grassland system in these zones play an important task of water conservation and water purification by adjusting the mountainous hydrological paths. Chosen forest-grassland system as study area, this study simulating the effects of forest-grassland system on water yields in mountain regions in order to provide a scientific basis for the establishment of a reasonable strategy of basin’s water management.This study chose Tianlaochi catchment as typical study area and using GIS and RS technologies to discretize the weather, vegetation and soil data. By using the discrete data as input data, this study used SWAT model to simulate the effect of vegetation change on runoff generation during the rainy season. The main conclusions of this study are as follows:The optimal parameter set for GcoEye-l’s image segmentation is [120,0.2,0.4]. Using the segment parameter set, the overall accuracy of the classification result is90.25%and the total KAPPA coefficient is0.86.Area of the zone which is range from2800m to3800m accounts for about85%of area of Tianlaochi catchment and the zone is dominated by forest-grassland system. Bare rocks are mainly distributed above the altitude of3800m. shady area occupied the largest area in Tianlaochi catchment, followed by the semi-shady area. These two areas account for61.44%the total area while sunny and semiarid sunny areas accounting for only38.56%of the total area. Sabina przewalskii, desert steppe and alpine meadow are mainly distributed on sunny and semi-sunny area while shrubs and picea classifolia area mainly distributed on shady and semi-shady area.The variability of soil organic matter content, soil bulk density and porosity that was caused by the variability of vegetation and other environmental factors can result in different soil hydrological characteristics, and thus affect water holding capacity and water storage capacity. Even if the same soil type, the different soil organic matter content, soil bulk density, soil texture and thickness of the litter layer could also lead to different saturated hydraulic conductivity and soil water retention capacity in(the roots zone, which result in different hydrological processes in soil layers.Because of the advantage of considering different types of environmental variables as explanatory variables in classification and regression tree model (CART) and generalized linear models (GLM), these models become the preferred model for spatial mapping of soil data. However, it should be noted that the development of reliable variograms is difficult for the Tianlaochi catchment where sampling intensity is low and a range of soil-landscape exist with contrasting pedogenic processes and presumably different scales and patterns of spatial dependence.The measured data has an important influence for parameter sensitivity analysis in the SWAT model, so the measured data should be used in the model parameter adjustment. Except the26parameters in sensitivity analysis, lateral flow travel time (LATTIME) and slope length for lateral subsurface flow (SLSOIL) are also sensitive for the simulated results. After calibrating, the best parameters for runoff simulation in SWAT model are:baseflow alpha factor (ALPHABF) is0.065, initial SCS runoff curve number for moisture Ⅱ (CN2) is added to-10, soil evaporation compensation (ESCO) factor is0.95, effective hydraulic conductivity in main channel alluvium (CH_K2) is20, saturated hydraulic conductivity is (SOL_K) added10, threshold depth of water in the shallow aquifer required for return flow (GWQMN) to occur is1, average slope of subbasin is0.75(SLOPE) and available water capacity of the soil layer (SOL AWC) is added to-0.04.Alter calibrating the model parameters use observed data, the Nash-Suttcliffe coefficient reached0.67and coefficient of determination reaches0.71during calibration stage. It indicates that SWAT model does well in simulating the rainfall process in Tianlaochi catchment. In the case of no rainfall or small rainfall occurred, the simulating results are better, and after a long duration of rainfall case the simulating results are poor. When arboreal forest transit into shrubs and alpine meadow, the catchment water yield will increase the rates of6.38%and6.65%for the current vegetation cover. When the shrubs which are below3400asl transit into arboreal forest, the catchemnet water yield will decrease the rates of9.28%. According to the simulating results from SWAT model, we conducted that for a10%change of catchment area:1arboreal forest transit into shrubs, water yield increased1.20mm;2shrubs transit into alpine meadow, water yield increasing0.13mm; shrubs transit arboreal forest, water yield increased2.58mm. Surface runoff account for only a very small proportion of the total water yield. The surface runoff generated in forest-grassland system mainly generated by alpine meadow, and the surface runoff generated in forests is less than which generated in alpine meadow. Because of the high degree of vegetation cover, vegetation conversion from arboreal forest to shrubs and alpine meadow, the increase in water yield comes mainly from subsurface flow and ground flow. |
PDF Full Text Request