陆面过程模型CoLM与BATS1e的模拟精度比较——以东北玉米农田为例

Comparison of Land Surface Process Models Performance Between CoLM and BATS1e: A Case Study on Maize Field in Northeast China

利用2008年锦州玉米农田生态系统野外观测站资料对CoLM与BATS1e模型模拟能力进行定量评价。比较发现:两模型对净辐射和表层土壤温度的模拟精度都较高且差异不大,CoLM模型对感热、潜热、土壤热通量、次表层土壤温度模拟能力都不同程度高于BATS1e模型,模拟值对实测值解释能力分别偏高3%、22%、1%、10%,NS(Nash-sutcliffe效率系数)分别偏高1.042、0.266、0.023、0.138。从各月情况看,两模型在7月对感热、潜热模拟能力都较高,而在其它月份CoLM模型模拟精度明显高于BATS1e模型,土壤热通量和次表层土壤温度在5~8月前者模拟精度高于后者。由于CoLM模型对潜热更高精度的模拟可证明其对表层土壤湿度模拟精度高于BATS1e模型。

Using observations data in 2008 from Jinzhou Agricultural Ecosystem Research Station, land surface process models performance between CoLM and BATS1e were evaluated and compared. The results show that net radiation and surface soil temperature are both able to be accurately simulated with a very small difference between the two models. Model performance of CoLM is higher in different degrees than BATS1e for simulations of sensible, latent, soil heat fluxes and subsurface soil temperature with interpreting ability of simulations to observations and Nash-sutcliffe coefficient of efficiency factor of the former being 3%,22%,1%,10%and 1.042,0.266,0.023,0.138 higher than those of the later, respectively. As far as every month is concerned,simulation precisions of sensible and latent heat fluxes in most months except July when two models work very well as well as soil heat fluxes and subsurface soil temperature from May to August are higher for CoLM than BATS1e. Besides, it will be proved that Model performance of CoLM for simulation of surface soil water content is also higher than that of BATS1e because of its more accurate simulation for latent heat flux.