The application of Rapid Scan data to the Convective Rainfall Rate algorithm from SAF NWC for the area of the Czech Republic

0426394 - ÚFA 2015 RIV NL eng J - Článek v odborném periodiku
Bližňák, Vojtěch - Sokol, Zbyněk - Pešice, Petr
The application of Rapid Scan data to the Convective Rainfall Rate algorithm from SAF NWC for the area of the Czech Republic.
Atmospheric Research. Roč. 144, July 2014 (2014), s. 82-94. ISSN 0169-8095. E-ISSN 1873-2895
Grant CEP: GA MŠMT LD11044; GA MŠMT ME09033; GA ČR GA205/07/0905
Institucionální podpora: RVO:68378289
Klíčová slova: meteorological satellite * weather radar * convective storm * satellite precipitation estimates * CRR algorithm * Czech Republic
Kód oboru RIV: DG - Vědy o atmosféře, meteorologie
Impakt faktor: 2.844, rok: 2014
http://www.sciencedirect.com/science/article/pii/S0169809512002724

The goal of this paper is to calculate new calibration matrices using Rapid Scan (RSS) Meteosat Second Generation (MSG) measurements and to evaluate their impact on precipitation estimates for a territory of the Czech Republic. The calibration matrices are the most important part of the Convective Rainfall Rate algorithm, which uses the IR 10.8 μm, WV 6.2 μm and VIS 0.6 μm spectral SEVIRI (Spinning Enhanced Visible and Infrared Imager) channels of the MSG to assess satellite precipitation estimates (SPEs). The calibration matrices were calculated using Czech radar data from 21 summer days during which severe convection and heavy precipitation were observed. The resultant matrices were compared with those obtained using conventional 15 min MSG scans. The comparison showed significant differences in the calibration matrices, which resulted in differences in the estimated precipitation. The application of RSS data significantly increased the rain rates and improved the structure of the matrices; however, the matrices were subjectively modified to increase the accuracy of the resulting SPEs. The calibration matrices were also calibrated by shifting the radar data forward 5, 10, 15 and 20 min with respect to the MSG measurement, because some delay between the information obtained by the MSG and the radars was expected. The impact of the matrices was evaluated by verifying the SPEs with the radar-derived precipitation estimates merged with the rain gauge observations as the ground truth. The results showed that the calibration matrices that were based on the RSS data improved the categorical skill scores and reduced the mean error (ME), the mean absolute error (MAE), and root mean square error (RMSE) of SPEs.
Trvalý link: http://hdl.handle.net/11104/0232087