Operational Rainfall Retrieval Based on Cloud Microphysical Properties

A modular rainfall retrieval called Advective-Convective-Technique (ACT) was developed at the LCRS that is based on data from geostationary sensors in the water vapour and infrared channel. Extended regions of stratiform precipitation in the vicinity of convective cores are identified by a kmeans cluster analysis and rainrates are assigned using a 3-D cloud model. During daytime an extended identification scheme (CP-ACT) with respect to data in the near-infrared and visible spectrum can be used for a more reliable identification of cloud regions with advective/stratiform precipitation processes by taking into account the cloud optical thickness and effective droplet radius. The cloud parameters are retrieved by a semi-analytical approach using asymptotic radiative transfer equations which is almost as accurate as the common look-up table approaches but much less time consuming. Within the framework of the GLOWA-Danube project the ACT has been applied to half-hourly Meteosat VISSR data of the upper Danube catchment area covering the alpine forelands of southern Germany as well as the Austrian and Italian alps. Evaluations against data from 45x45km MM5 model runs as well as against interpolated point measurements from about 250 stations in this area show good agreement. Since there have been some calibration problems regarding the Meteosat-8 SEVIRI channels the CP-ACT module was developed using data from the Terra-/Aqua-MODIS systems. Comparisons against ground based radar data from the German weather service shows good agreement. 1. THE ADVECTIVE-CONVECTIVE-TECHNIQUE ACT Operational infrared/water vapour channel rainfall retrievals predominantly focus on tropical/subtropical regions. The straightforward convective schemes which normally identify potential precipitating clouds by means of their infrared brightness temperature usually perform well in the tropics but cannot simply be applied to the complex situation of mid-latitude frontal precipitation. Hence, a new modular retrieval scheme, the Advective-Convective-Technique (ACT) that is also applicable to advective precipitation in the mid-latitudes has been developed. It consist of three modules. The first deals with precipitation from convective core areas, the second from that of advective cloud regions and the third is an enhanced classification scheme especially useful for stratiform precipitating clouds using additional information of cloud properties, namely the cloud optical thickness τ and the effective droplet radius aef. Because the first two modules only require brightness-temperatures from the infrared (TBIR) and water vapour (TBWV) channels, 2 Workshop of the International Precipitation Working Group they can be used to investigate existing long time series of geostationary data (e.g. GOES, Meteosat). However, the increased spectral resolution of the latest generation of geostationary satellites (especially spectral bands at 0.6 and 1.6μm) is necessary for the third module (CP-ACT). Figure 1 presents the principal outline of the ACT scheme. The ACT convective module is based on the Enhanced Convective Stratiform Technique (ECST, Reudenbach et al. 2001, Reudenbach 2003) that uses positive TBWV TBIR differences (dWI) in order to discriminate between deep convective, optically thick clouds (dWI>0) and non-raining cirrus (dWI<0, refer to Tjemkes et al. 1997). Pixels with positive dWI are then subdivided by analysing the frequency distribution of TBIR. Areas with TBIR<1 quartile of the frequency distribution represent overshooting tops of convective cores, those who suit the 1 quartile reveal raining systems at tropopause level and pixels with TBIR <3 quartile identify potentially raining cloud systems of high vertical extension. As a result, isolated convective cores can be distinguished from directly adjacent stratiform raining areas.