The ongoing NASA/JAXA Global Precipitation Measuring mission (GPM) requires the full exploitation of the complete constellation of passive microwave (PMW) radiometers orbiting around the globe for global precipitation monitoring. In this context the coherence of the estimates of precipitation using different passive microwave radiometers is a crucial need. Within the EUMETSAT H-SAF program (Satellite Application Facility on Support to Operational Hydrology and Water Management, http://hsaf.meteoam.it) we have developed two different passive microwave precipitation retrieval algorithms optimized for the MSG full disk area, with focus on Europe, the Mediterranean area, Africa and Southern Atlantic. One is the Cloud Dynamics Radiation Database algorithm (CDRD), founded on a physically-based Bayesian approach for conically scanning radiometers (i.e., DMSP SSMIS); the other one is the Passive microwave Neural network Precipitation Retrieval (PNPR) algorithm for cross-track scanning radiometers (i.e., NOAA and MetOp-A/B AMSU-A/MHS). In order to achieve the coherence in the estimates, the two algorithms are based on the same physical foundation, i.e., same cloud-radiation model simulations to be used as a priori information in the Bayesian solver and as training dataset in the neural network approach. Both algorithms use dynamical/me teorological/environmental variables as ancillary information to characterize the observed event, and mitigate the ambiguity of the cloud microphysical structures (and rainfall rates at the ground) associated to any given set of measured multichannel brightness temperatures. The cloud-radiation database which is the physical base for both algorithms is built upon a large number of cloud resolving model simulations of the different types of precipitation events covering the different climatic regions in the area of interest, and taking into account its variable land surface conditions. Two years of SSMIS and AMSU/MHS data have been considered to carry out a verification study over Africa of the retrievals from both algorithms. The precipitation estimates from the TRMM-Precipitation radar (PR) (TRMM product 2A25) have been used as ground truth. The results of this study aimed at assessing the accuracy of the precipitation retrievals in different climatic regions and precipitation regimes will be presented. Particular emphasis will be placed to the analysis of the level of coherence of the estimates between the two PMW algorithms. In order to enhance the consistency between the precipitation patterns estimated by the CDRD and PNPR algorithm a novel algorithm for the detection of precipitation in tropical areas has been developed and tested. The novel detection algorithm is applicable to every PMW radiometer and shows a small rate of false alarms and a superior detection capability in comparison with many widely used comparable algorithms.
CDRD and PNPR passive microwave precipitation retrieval algorithms: extension to the MSG full disk area
D Casella;G Panegrossi;P Sanò;S Dietrich;L Milani;M Petracca;
2014
Abstract
The ongoing NASA/JAXA Global Precipitation Measuring mission (GPM) requires the full exploitation of the complete constellation of passive microwave (PMW) radiometers orbiting around the globe for global precipitation monitoring. In this context the coherence of the estimates of precipitation using different passive microwave radiometers is a crucial need. Within the EUMETSAT H-SAF program (Satellite Application Facility on Support to Operational Hydrology and Water Management, http://hsaf.meteoam.it) we have developed two different passive microwave precipitation retrieval algorithms optimized for the MSG full disk area, with focus on Europe, the Mediterranean area, Africa and Southern Atlantic. One is the Cloud Dynamics Radiation Database algorithm (CDRD), founded on a physically-based Bayesian approach for conically scanning radiometers (i.e., DMSP SSMIS); the other one is the Passive microwave Neural network Precipitation Retrieval (PNPR) algorithm for cross-track scanning radiometers (i.e., NOAA and MetOp-A/B AMSU-A/MHS). In order to achieve the coherence in the estimates, the two algorithms are based on the same physical foundation, i.e., same cloud-radiation model simulations to be used as a priori information in the Bayesian solver and as training dataset in the neural network approach. Both algorithms use dynamical/me teorological/environmental variables as ancillary information to characterize the observed event, and mitigate the ambiguity of the cloud microphysical structures (and rainfall rates at the ground) associated to any given set of measured multichannel brightness temperatures. The cloud-radiation database which is the physical base for both algorithms is built upon a large number of cloud resolving model simulations of the different types of precipitation events covering the different climatic regions in the area of interest, and taking into account its variable land surface conditions. Two years of SSMIS and AMSU/MHS data have been considered to carry out a verification study over Africa of the retrievals from both algorithms. The precipitation estimates from the TRMM-Precipitation radar (PR) (TRMM product 2A25) have been used as ground truth. The results of this study aimed at assessing the accuracy of the precipitation retrievals in different climatic regions and precipitation regimes will be presented. Particular emphasis will be placed to the analysis of the level of coherence of the estimates between the two PMW algorithms. In order to enhance the consistency between the precipitation patterns estimated by the CDRD and PNPR algorithm a novel algorithm for the detection of precipitation in tropical areas has been developed and tested. The novel detection algorithm is applicable to every PMW radiometer and shows a small rate of false alarms and a superior detection capability in comparison with many widely used comparable algorithms.I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.
