基于三维模型ENVI-met对黑河森林和北方森林的潜热及显热通量模拟

Modeling sensible flux and latent flux in Heihe and boreal forests based on a 3D ENVI-met model

在非均匀混合场景下模拟区域能量,三维小气候模型比一维模型考虑因素更为全面,但三维模型较复杂,研究甚少。为探究三维模型在混合像元场景中的通量模拟能力,以ENVI-met模型为例,以黑河关滩森林和加拿大北方森林为实验区,比较森林通量模拟值和实测值的季节性和日变化差异,并模拟森林三维空间结构对通量分布的敏感性。结果表明:(1)模型在春季、夏季、秋季模拟拟合度(R^2)分别为0.75,0.76,0.55。模拟显热通量、潜热通量日变化符合实测趋势,日平均实测值与模拟值之差均不超过10.0%。模型对显热通量、潜热通量和太阳短波辐射模拟精度分别为0.68,0.63,0.80。(2)模拟总体能量闭合率(EBR)达92.9%,模型对气象参数微小变动不敏感,较为稳定。(3)自然参数不变场景下,团状分布通量值大于均匀分布,变化幅度剧烈,且显热通量对空间异质性较潜热通量更为敏感。研究结果肯定了ENVI-met模型模拟通量具有较好精度,且能够反映出森林空间分布格局在通量分配上的影响,为研究区域气候变化提供了有效手段。

Three-dimensional microclimate models within a heterogeneous mixed scene consider more structural factors than a one-dimensional, however, a 3D model has relatively few reports because of its complexity theory. To evaluate the capacity and potential of 3D models in terms of flux energy simulation, this research used the ENVI-met model, as an example, with data from the Chinese Heihe program and the Canadian BOREAS pro-gram. Daily and seasonal variation between observed and simulated results, as well as model sensitivity in forests with different spatial distribution patterns, was analyzed. Results （1） for fitted R^2 values of the model showed that in spring R^2=0.75, in summer R^2=0.76, and in autumn R^2=0.55. Modeling day change tendency was consistent with observed data where differences between observed and simulated daily sensible and latent heat fluxes were below 10%. Accuracy for sensible heat flux, latent heat flux and solar shortwave radiation simula-tion were 68%, 63% and 80%, respectively. （2） The model＇s energy balance ratio （EBR） was 92.9%, meaning it was relatively stable and insensitive to meteorological parameters. （3） The sensible heat flux value of a lumped forest distribution was 155.52 W·m^-2 higher than an uniform distribution with daily changes in a wide range. Overall, ENVI-met had favorable accuracy in terms of surface energy flux simulation, and reflected flux energy distribution differences caused by forest spatial distribution patterns; so the 3D model could provide an effective means to regional climatology research.