A REGIONAL CLIMATE MODEL FOR THE WESTERN UNITED-STATES
A REGIONAL CLIMATE MODEL FOR THE WESTERN UNITED-STATES
复制标题
DOI：
10.1007/bf00240465
复制
发表时间：
1989-12-01
期刊：
影响因子：
4.8
通讯作者：
BATES, GT
中科院分区：
文献类型：
--
作者：
DICKINSON, RE;ERRICO, RM;BATES, GT
一个数值方法来模拟气候在区域尺度上的发展，从而大尺度天气系统模拟与全球气候模式（GCM）和GCM输出是用来提供高分辨率的中尺度模式模拟所需的边界条件在该地区的利益。在我们的例子中，我们使用美国国家大气研究中心（NCAR）社区气候模式（CCM 1）和宾夕法尼亚州立大学（PSU）/NCAR中尺度模式第4版（MM 4）在美国西部应用这种方法。由CCM 1的500公里网格所分辨的地形必然是高度扭曲的，但使用MM 4的60公里网格，主要山脉是可以区分的。为了获得足够的和一致的地表气候表示，我们使用相同的辐射和陆地表面处理在这两个模型中，后者是最近开发的生物圈-大气传输计划（BATS）。我们的分析强调了在四个CCM 1点周围的尤卡山，内华达州，因为需要确定其气候认证之前，作为一个高水平的核废料reposition.We模拟全球气候三年的CCM 1/BATS和描述所产生的1月的表面气候在美国西部的降水模式的细节是不切实际的，因为平滑的地形。选择5个1月的CCM 1风暴发生在美国西部，总持续时间为20天的模拟与MM 4，我们证明了中尺度模型提供了更好的冬季降水模式。MM 4中的风暴单独比CCM 1中的风暴更加真实。一个简单的平均程序，推断平均1月降雨气候计算从20天的MM 4模拟比CCM 1气候更接近观测。然而，在MM 4的3-5天积分期内模拟的土壤水分和地下排水仍然强烈依赖于初始CCM 1土壤水分，因此比降雨更不现实。地表土壤水的充分模拟可能需要中尺度模型在一段时间内的整合。
A numerical approach to modeling climate on a regional scale is developed whereby large-scale weather systems are simulated with a global climate model (GCM) and the GCM output is used to provide the boundary conditions needed for high-resolution mesoscale model simulations over the region of interest. In our example, we use the National Center for Atmospheric Research (NCAR) community climate model (CCM1) and the Pennsylvania State University (PSU)/NCAR Mesoscale Model version 4 (MM4) to apply this approach over the western United States (U.S.). The topography, as resolved by the 500-km mesh of the CCM1, is necessarily highly distorted, but with the 60-km mesh of the MM4 the major mountain ranges are distinguished. To obtain adequate and consistent representations of surface climate, we use the same radiation and land surface treatments in both models, the latter being the recently developed Biosphere-Atmosphere Transfer Scheme (BATS). Our analysis emphasizes the simulation at four CCM1 points surrounding Yucca Mountain, NV, because of the need to determine its climatology prior to certification as a high-level nuclear waste repository.We simulate global climate for three years with CCM1/BATS and describe the resulting January surface climatology over the western U.S. The details of the precipitation patterns are unrealistic because of the smooth topography. Selecting five January CCM1 storms that occur over the western U.S. with a total duration of 20 days for simulation with the MM4, we demonstrate that the mesoscale model provides much improved wintertime precipitation patterns. The storms in MM4 are individually much more realistic than those in CCM1. A simple averaging procedure that infers a mean January rainfall climatology calculated from the 20 days of MM4 simulation is much closer to the observed than is the CCM1 climatology. The soil moisture and subsurface drainage simulated over 3–5 day integration periods of MM4, however, remain strongly dependent on the initial CCM1 soil moisture and thus are less realistic than the rainfall. Adequate simulation of surface soil water may require integrations of the mesoscale model over time periods.