<title>Comparison and integration of active and passive microwave for rainfall monitoring: a study case</title>

Abstract

Firenze (Italy)°University of Nantes, IRESTELa Chantrerie CP 3003, 44087 Nantes Cedex 03 (France)ABSTRACTThis paper reports the experimentation of ajoint use of ground-based radar and satellite radiometer to calibrate andvalidate passive microwave data for rainfall monitoring. In particular, the utilisation of radar vertical profile maps inorder to figure out the passive microwave signatures is outlined. The experimentation has been conducted utilisingactive microwave data acquired from the Chilbolton radar and passive microwave data acquired from the DMSPSSMII instruments. The two microwave data type have been processed in order to be overlayable in time and space,and several parameters have been estimated from the SSM/I data. Radar vertical profile maps have been classified tooutline the cloud structure and the obtained information have been utiised to make statistical analysis of passivemicrowave parameters related to rainfall over land. In particular, an attempt to divide the cloud structure into threemeaningful layers has been worked out; this is important as the different channels of SSMII can penetrate in a diverseway cloud structure.The results are in accordance to theory and outline the importance of utilising radar vertical profiles in order to figureout the passive microwave measurements.Keywords: active and passive microwave integration; calibration and validation of passive microwave; rainfallmonitoring over land; active microwave; passive microwave; cloud structure classification.1. INTRODUCTIONSatellite passive microwave measurements of precipitation are characterised by the presence of large hydrometeorswithin clouds, they view the atmosphere from the top downwards and the spatial resolution is very poor. For rainfallmonitoring applications utilising satellite passive microwave, several important issues must be tackled: the beamfilling problem and the inversion problem as well as the spatial and temporal scarce resolution. Surface-based radarmeasurements of precipitation are based on the estimation of hydrometeors falling from the bases of clouds, they viewthe atmosphere from the bottom upwards way. The temporal and spatial resolution is very good but there are severalphysical and technological reasons why the radar estimates may be cumbersome1.Due to passive microwave capability to penetrate clouds and directly sense precipitation particles, much rainfallalgorithms development has been recently based on measurements in the passive microwave regime2. Two basicapproaches have been used to detect rain using passive microwaves: Regression based algorithms and PhysicalInversion algorithms. As far as complexity is concerned, regression methods are simple and relatively undemanding ofcomputer resources, but they are not very flexible. Physical inversion algorithms are based on coupled cloud-radiationmodels, in which the cloud model component is an explicit dynamic-microphysical cloud-mesoscale model. They aregenerally complex and are characterised by large computational requirements but they provide flexibility3. It is widelyaccepted that the most satisfactory in the mid-term future are likely to be physical/empirical methods based on goodknowledge of cloud physics and use of suitable cloud/rainfall models. Studies on rain events utilising data collected byboth radar and satellite are required; in particular, the utilisation of radar vertical profile information to guess cloudstructure must be particularly considered. The present work wants to be a contribution toward this direction.Investigation of the vertical profile in the presence of precipitation is one of the main topics of interest in radarmeteorology, for different kind of applications such as: micrometeorology and the detection and correction of effectdue to the presence of bright band in radar rainfall measurements for meteorological application4.