Study On Remote Sensing Model Construction And Estimation Of Water Production Efficiency In Weihe River Basin

Water Use Efficiency (WUE) can be used to describe the relationship between "water loss" and "carbon fixation" of plants in the process of photosynthesis, which is an important variable to link ecosystem carbon and water cycles and an important indicator of the coupling relationship between water and carbon cycles. Studying WUE can enhance our ability to reveal how global change affects water and carbon cycles. At regional scale, WUE can be calculated through the ratio of Net Primary Production (NPP) and evapotranspiration (ET).In the recent years, Wei River Basin suffers serious ecological degradation under the double impact of climate changes and human activities and it faces with the dual task of saving water resources and ecological restoration. Developing WUE remote sensing, estimating NPP and ET and studying the basic features of WUE in Wei River Basin can provide the basis for quantifying the effective of ecological protection measures and environmental management policies and for the integrated management and macro-control of water and carbon resources. What’s more, it also can provide a reference for weighing the relationship between ecological restoration and lacking water resources, in other words, carbon fixation and water consumption.Based on triangle model and CASA model which is used to estimate ET and NPP, this paper construct the WUE remote sensing models with improving some parameters. Using that model, this paper achieves the estimation of NPP by the scale of 16 days, ET by the scale of everyday and WUE by the scale of 16 days in Wei River Basin in 2010 based on related MODIS imagery and meteorological dates. Then the estimated results are compared, verified and analyzed. The main conclusions are as follows:(1) The result of NPP estimated by CASA model, which has been improved, is better than MOD17-NPP in Wei River Basin. The results of NPP estimated by different kinds of NPP models are different because the models are estimated in different year or by different data sources or by different parameters. Evaluate of cumulative NDVI show that the NPP model constructed by this paper is reasonable in stimulating the temporal and spatial distribution of NPP in Wei River Basin.(2) ET estimated by triangle model in Wei River Basin has good comparability with MOD16-ET in temporal and spatial distribution and good correlation with the result estimated by FAO Penman-Monteith equation in time series. The verification result of ET between triangle model based on remote data and water balance method based on meteorology and hydrology data show that the average relative error of two results about ET is 6.51%.(3) The result of WUE estimated by WUE remote sensing model constricted by this paper has a good correlation with MODIS-WUE in temporal and spatial distribution over Wei River Basin. The simulation results of WUE model is poor to evergreen coniferous forest and deciduous conifer forest because of the less pixel number, while the simulation results of other vegetation is relatively good and has a comparability with observational data of flux tower and the simulation results of ecosystem model. The results of WUE estimated by different kinds of WUE models are different because the models are estimated with different WUE definition, different simulation area, different data sources, or different vegetation classification map.(4) The monthly variation of NPP roughly shows a double peak pattern over Wei River Basin in 2010 with highest value in August. The seasonal variation of ET showed that the maximum value is in summer, then autumn, spring, winter. The monthly variation of ET roughly shows a unimodal pattern over Wei River Basin in 2010 with highest value in June. The seasonal variation of ET showed that the maximum value is in summer, then spring, autumn, winter. The spatial distribution of NPP and ET over Wei River Basin shows that the high value pixel is in forest region of Liupan Mountain, Ziwuling, Huanglong Mountain and northern slope of Qinling, while the low value pixel is in the low vegetation coverage district of upper basin and buildup region of Xi’an city.(5) The annual average variation of WUE showed that the maximum value is in forest ecosystem, then farmland ecosystem, grassland ecosystem, shrub ecosystem, over Wei River Basin in 2010. The monthly variation of WUE roughly shows a double peak pattern over Wei River Basin in 2010 with highest value in August. The seasonal WUE showed that the maximum value is in summer (1.05 gC·m-2·mm-1), then autumn (0.66 gC·m-2·mm-1), spring (0.57 gC·m-2·mm-1), winter (0.12 gC·m-2·mm-1). The overall spatial distribution of WUE shows that the maximum value is in the upstream of Wei River Basin, then midstream, downstream. The regional spatial distribution of WUE shows that the high value pixel is in forest region of Ziwuling, Huanglong Mountain, Liupan Mountain and northern slope of Qinling, while the low value pixel is in the build-up region of Xi’an city, low vegetation coverage district of upper basin and partly dry farming areas.(6) With the increase of temperature, the change of WUE over Wei River Basin can be divided into five stages, which is essential invariant (-10～-5℃), slight increase (-5-0℃), rapid increase (0-20℃), stabilization (20～25℃) and decline (>25℃). With the increase of precipitation, the change of WUE over Wei River Basin also can be divided into five stages, which is rapid increase (0-50 mm), slow increase (50～100 mm), stabilization (100～200 mm), slow decline (200～250 mm), rapid decline (>250 mm).