Combination of ground-based and satellite cloud observations on a routine basis

A Cloud Detection System (CDS) is designed, installed and operated for a two year period (1995/1996). In the ground-based network ten clouds and radiation stations are distributed over a 120 x 120 km(2) area. Every station consists of a lidar ceilometer, an m-radiometer and a pyranometer. The ground based observations are combined with the observations from the operational meteorological satellites Meteosat and NOAA/AVHRR. The CDS was intended to provide a data set for the evaluation of atmospheric model output. From an analysis of the data for a two month period (January/February 1995) it is shown that the synoptical cloud cover observations are reproduced with a standard deviation of two octa for a single station and one octa for the CDS-area averaged values. The CDS data allow for the derivation of the following cloud parameters: both height and temperatures of cloud base and top, optical depth and emissivity. In this paper an example is given of the derivation of the cloud emissivity in the spectral region of the IR-radiometers. In this article the set-up of the CDS is described, together with a description of the algorithms which are used to retrieve the cloud parameters. The usefulness of the CDS data is illustrated with several examples. The CDS data set contains a two years time series of objective cloud data that can be used for model evaluation and for the validation of satellite retrievals. The CDS concept is widely applicable. Worldwide there is a large number of potential ground based stations. This has not been fully exploited yet.