Initial cloud detection by comparing brightness temperatures derived from a numerical weather prediction model with those observed

Abstract

A cloud detection scheme is being developed at the UK Met. Office that compares the observed satellite sounder brightness temperatures calculated from a short term forecast of a numerical weather prediction (NWP) model. The scheme will be used in the first step of the operational processing of sounder data for assimilation into the NWP models. The data are from TOVS on the NOAA polar orbiting satellites, and will soon also include data from the new advanced TOVS (ATVOS). The aim of the scheme is to use one calculation to maximize the information assimilated from the infrared channels into the NWP model, while preventing the assimilation of cloud contaminated data. Where no cloud is detected by the scheme the maximum number of infrared channels shall be used in the subsequent retrieval of the data. Where cloud is detected by the scheme the affected infrared channels will be excluded from the later 1DVAR processing. Being in the same program as the 1DVAR, the existing error covariance matrices and radiative transfer code are readily available to the cloud detection scheme. Details are given of the principles behind the cloud detection scheme. The cloud detection results are compared with data form: AVHRR imagery, the results from a 1DVAR cloud retrieval and the current operational cloud detection scheme.© (1996) COPYRIGHT SPIE--The International Society for Optical Engineering. Downloading of the abstract is permitted for personal use only.