The Relationship Between Cloud Vertical Overlap And Cloud Physical Properties And Wind Shear Intensity

Based on the observation of the Cloud Sat / CALIPSO satellite from 2007 to 2010,the key parameter used to express the overlapping relationship in the vertical direction of the cloud layer in the climate model——the decorrelation length(L)is calculated.The relationship between L and the cloud fraction profiles categorized by self-organizing maps(SOM)algorithm,cloud optical thickness as well as cloud types are analyzed.The seasonal variation characteristics of L and wind shear intensity are further studied.The main conclusions are given as follows:(1)When the total cloud amount and the cloud optical thickness for the entire atmosphere column are small,the clouds tend to overlap randomly which lends small L.For the cloud with similar cloud fraction profiles,L is generally larger for the ones with obvious peaks in the cloud fraction profiles.In terms of cloud types,deep convective clouds generally contribute significantly to the value of L.The structure of the nimbostratus tends to be more inhomogeneous,leading to a very small L generally.It shows that different cloud types are significantly correlated with L.We also find that L exhibited a remarkable single-peaked variation with height for most of the globe.(2)It is found that L has an obvious negative correlation with wind shear on a global scale,indicating that clouds at different heights tend to overlap randomly in the vertical direction with the increase of wind shear.The relationship between L and wind shear is more significant in the northern hemisphere's middle latitudes,and the correlation coefficient reaches-0.94.However,the negative correlation between L and wind shear intensity is significantly weaker at middle and high latitudes in the southern hemisphere,where the ocean surface is dominant,because of the long-term presence of deep convective cloud systems in this latitude region and the weaker influence of meteorological conditions on the overlapping relationship in the vertical direction of clouds.On this basis,a parametric relationship between the decorrelation length scale L and wind shear intensity is established.Compared with other parametric schemes,the new scheme in this paper can effectively improve the simulation of cloud cover.(3)The spatial sampling scale of clouds used to calculate L shows an approximately linear positive correlation with the corresponding L of cloud samples.It indicates that clouds of different heights tend to overlap more in the vertical direction as the spatial sampling scale of clouds becomes larger.In addition,the coefficients in the parametric relationship between L and wind shear intensity also differ under different spatial sampling scales.Therefore,the spatial sampling scale of clouds is also an important factor to consider when establishing a refined L parameterization scheme in climate models.