模型驱动数据空间分辨率对模拟生态水文过程的影响

设定两种独立的调参方案,检验将高/低分辨率下的调参值应用到低/高分辨率中驱动模型模拟的可行性;同时分析驱动数据分辨率的差异对模型模拟生态水文过程的影响;此外,设置9种集水面积阈值,分析驱动数据分辨率与集水面积阈值对模拟结果的影响。结果表明两种调参方案下模拟的水文与生态数据相关性强、差异小,说明调参方案受驱动数据分辨率的影响不明显;分辨率对土地覆被和土壤类型面积影响小,而对高程和坡度影响大;驱动数据分辨率降低,模拟径流的均值减小、峰值增大、生态变量值增大;驱动数据分辨与集水面积阈值之间不存在交互作用;集水面积阈值增大,模拟径流的均值、峰值均减小,生态变量值也减小,但阈值变化对峰现时间无影响。

Effects of spatial resolution of model-driving data on simulating land-surface eco-hydrological processes

This study used CHESS (The Coupled Hydrology and Ecology Simulation System) to explore (1) how the model's parametrization scenarios affect model simulation driven by forcing data of different spatial resolutions; (ii) how the spatial resolutions of model-forcing data affect model results; and (iii) how the selection of drainage area threshold affects model simulations. Eight sets of driving data of different spatial resolutions, two independent parameterization schemes A and B as well as nine sets of drainage area thresholds are developed in this study for analyses. As for the parameterization scenarios, the scheme A/B first parameterizes the model based on high-resolution/low-resolution forcing data. Then, the parameterization results are applied to simulations driven by low-resolution/high-resolution data. Based on the results from the parameterization scenario B, the effects of spatial resolution of model driving data on simulated land-surface eco-hydrological processes are investigated. Our results indicate that the simulated eco-hydrological variables between schemes A and B are strongly correlated and their differences are minimal. This indicates that the model's parameterization is not sensitive to the resolutions of model driving data. Additionally, the resolution of model driving data has little effect on the land cover and soil types but it greatly affects the elevation and slope of the study catchment. With the decrease in spatial resolution of model driving data, the elevation and slope of the catchment decrease, the micro terrains suffer from loss, and the terrain becomes flatter. Through the statistical analysis of simulated eco-hydrological variables, we found that the mean value of simulated runoff decreases, the peak value increases, and the value of ecological variable increases with the decrease in the resolution of the driving data. Besides, there is no significant interaction between the resolution of driving data and the selection of drainage area thresholds. With the increase in drainage area threshold, the mean and peak values of runoff decrease and so do the magnitudes of modeled ecological variables. Nevertheless, the selection of drainage area thresholds does not affect the peak time of runoff.