Effects Of Cloud Inhomogeneities On Remote Sensing Of Cloud Properties

Cloud is the major contributor to modulating the earth's radiation budget. It thereby plays a key role in the weather and globe climate. At present, the situation of the cloud-radiation feedback varies greatly between different General Circulation Models (GCMs), and even the sigh of the feedback is uncertain. Uncertainties such as this have motivated research in clouds and their radiative properties, and satellite remote sensing of clouds is expected to improve understanding these uncertainties.Cloud remote sensing techniques are generally based on plane-parallel theory and Independent Pixel Approximation (IPA) method, in which the horizontal radiative transfer between pixels is neglected. It may cause errors in actual clouds which possess horizontal inhomogeneity. Some work addressed the bias in retrieved cloud parameters due to the cloud horizontal inhomogeneity, but it is not well understood at present. More studies on this area should be taken.The present study aims at contribution to investigating the bias in retrieved cloud optical depth due to the cloud horizontal inhomogeneity in solar band. In order to simulate various cases of inhomogeneous cloud, we created a stochastic cloud model and employed 3D radiative transfer models to fit our aim. The retrieved biases of cloud optical depth are described for different cloud cases.As a general summary, it is suggested that the local standard deviation of optical depth is useful in revealing the local inhomogeneities of cloud field. It is found that IPA retrievals of optical depth look smoother than the true situation with almost overhead sun, while it look smoother or rougher with oblique sun, especially in the transparent waveband.