冰晶粒子不同形状假定对辐射收支和气候的影响

将包含多形状冰晶粒子的冰云辐射参数化方案应用于全球气候模式中，详细讨论了冰云粒子从球形假定到多形状假定的变化对辐射场和气候场的影响。结果显示，冰晶粒子形状假定的引入对冰云光学厚度、辐射通量和加热率以及温度场均有明显的影响。采用新的冰云方案使得全球平均云光学厚度值降低0.28（23%）；热带地区降低最为明显，其差异绝对值可达1.02，而在中高纬度陆地地区，两者的冰云光学厚度差别较小。冰晶粒子形状假定改变将导致全球平均的大气顶出射长波辐射通量增加5.52 W/m2（2.3%）。与观测资料的比较表明，多形状冰晶粒子假定明显减小了球形粒子假定对长波出射辐射的低估。对大气加热率廓线的模拟显示，多形状冰晶粒子假定会减弱短波辐射对大气的加热作用，同时增强长波辐射对大气的冷却作用；在热带对流层中高层，这两种影响尤为显著。冰晶粒子形状假定的改变对温度场有明显的影响，热带地区的对流层高层大气温度降低幅度可超过1.5 K。研究表明，冰晶粒子形状假定的改变对模拟的辐射和温度场均有重要的影响。 

Effects of different ice crystal shape assumptions on radiation budget and climate

Using a cloud radiative parameterization scheme with different assumptions of ice crystal shape in the global climate model, changes in radiation and climate between the spherical ice crystal assumption and the multi-shape ice crystal assumption are discussed in detail. The results show that the influences caused by ice crystal shape assumption on ice cloud optical depth, radiative fluxes, atmospheric heating rate and temperature are obvious. Compared with the spherical ice crystal assumption, the global average ice cloud optical depth decreases by 0.28 (23%) and the decrease is most obvious in the tropical area with the absolute value up to 1.02. Little changes are found in land areas in the middle and high latitudes. Global average upward longwave radiation flux at the top of atmosphere increases by 5.52 W/m² (2.3%). Compared with observations, the multi-shape ice crystal assumption significantly reduces the underestimation by the spherical ice crystal assumption. The multi-shape ice crystal assumption weakens the shortwave radiation heating effect but enhances the longwave radiation cooling effect in the atmosphere, and these effects are particularly significant in the tropical upper troposphere. The change in the ice crystal shape has a significant effect on temperature, and temperature in the tropical upper troposphere decreases by more than 1.5 K. In summary, the change in ice crystal shape assumption has an important influence on the simulated radiation and temperature.