Radiation and cloud radiative properties in the European Centre for Medium Range Weather Forecasts forecasting system
Radiation and cloud radiative properties in the European Centre for Medium Range Weather Forecasts forecasting system
复制标题
DOI：
10.1029/89jd01597
复制
发表时间：
1991-05
影响因子：
--
通讯作者：
J. Morcrette
中科院分区：
文献类型：
--
作者：
J. Morcrette
欧洲中期天气预报中心（ECMWF）业务模式中使用的辐射方案的连续版本（EC 1，EC2和EC 3），包括1989年5月2日开始运作的版本，进行了审查，并将其结果进行了比较，更详细的辐射模式的结果，由于在气候模式中使用的辐射代码的相互比较（ICRCCM）计划。对于晴朗天空条件，短波H2O吸收率在EC 1-EC2中被高估，这导致太大的短波大气吸收（高达20%）和太小的地面向下短波辐射（5-10%）。EC 1和EC2都明显低估了长波辐射冷却，主要误差在对流层低层，EC2在700和300 hPa之间。因此，EC 1和EC2在较小程度上低估了大气顶部的长波辐射。在多云条件下，EC 1显示出对少量散射体的夸大敏感性，这是EC2中纠正的问题。由于EC 1-EC2中嵌入了一个不切实际的模式云，这些方案不能正确地代表实际云液态水含量的短波行星辐射和出射长波辐射。EC 3修正了大多数这些缺陷，并给出了更好的协议与更详细的模型的结果。
The successive versions (EC1, EC2, and EC3) of the radiation scheme used in the European Centre for Medium Range Weather Forecasts (ECMWF) operational model, including the version which became operational on May 2, 1989, are reviewed and their results are compared to results of more detailed radiation models made available thanks to the Intercomparison of Radiation Codes Used in Climate Models (ICRCCM) program. For clear-sky conditions, the shortwave H2O absorptivity is overestimated in EC1-EC2, which leads to too large shortwave atmospheric absorption (by up to 20%) and too small downward shortwave radiation at the surface (by 5–10%). EC1 and EC2 both significantly underestimate the longwave radiative cooling, with main errors in the lower troposphere with EC1, and between 700 and 300 hPa with EC2. Therefore EC1 and, to a smaller extent, EC2 underestimate the outgoing longwave radiation at the top of the atmosphere. In cloudy conditions, EC1 shows an exaggerated sensitivity to small amounts of scatterer, a problem corrected in EC2. Due an unrealistic model cloud embedded in EC1-EC2, these schemes cannot properly represent both the shortwave planetary albedo and outgoing longwave radiation for realistic cloud liquid water contents. EC3 corrects most of these deficiencies and gives results in better agreement with those of more detailed models.