The New French Operational Radar Rainfall Product. Part II: Validation

Abstract

A new operational radar-based rainfall product has been developed at Météo-France and is currently being deployed within the French operational network. The new quantitative precipitation estimation (QPE) product is based entirely on radar data and includes a series of modules aimed at correcting for ground clutter, partial beam blocking, and vertical profile of reflectivity (VPR) effects, as well as the nonsimultaneity of radar measurements. The surface rainfall estimation is computed as a weighted mean of the corrected tilts. In addition to the final QPE, a map of quality indexes is systematically generated. This paper is devoted to the validation of the new radar QPE. The VPR identification module has been specifically validated by analyzing 489 precipitation events observed over 1 yr by a representative eight-radar subset of the network. The conceptual model of VPR used in the QPE processing chain is shown to be relevant. A climatology of the three shape parameters of the conceptual VPR (brightband peak, brightband thickness, and upper-level decreasing rate) is established and the radar-derived freezing-level heights are shown to be in good agreement with radiosonde data. A total of 27 precipitation events observed by three S-band radars of the network during the winter of 2005 and the autumns of 2002 and 2003 are used to compare the new radar QPE to the old one. Results are stratified according to the distance to the radar and according to the height of the freezing level. The Nash criterion is increased from 0.23 to 0.62 at close range (below 50 km) and from 0.35 to 0.42 at long range (between 100 and 150 km). The relevance of the proposed quality indexes is assessed by examining their statistical relationship with long-term radar–rain gauge statistics. Mosaics of QPE and quality indexes are also illustrated.

Review of the Different Sources of Uncertainty in Single Polarization Radar-Based Estimates of Rainfall

“…In the case of interception of the bright band, reflectivity can be typically overestimated by a factor of 2-5, up to a factor of 10 (e.g., Smith 1986;Joss and Waldvogel 1990;Fabry et al 1992;Bourrel et al 1994;Tabary 2007;Tabary et al 2007). Therefore, it is important to identify and correct for these regions (e.g., White et al 2002;Gourley and Calvert 2003;Mittermaier and Illingworth 2003).…”

Local influence of south-east France topography and land cover on the distribution and characteristics of intense rainfall cells

“…The data gathered from this radar were subjected to reliability analyses, with satisfactory results since 2001. More precisely, this radar data gathers eight successive processing stages (Tabary 2007;Tabary et al 2007;Champeaux et al 2009). These steps are a dynamic identification of ground clutter, a reflectivity to rain rate conversion using the Marshall-Palmer Z-R relationship (as mentioned above), a correction for partial beam blocking, a correction for vertical profiles of reflectivity (VPR) effect, a correction for non-simultaneity of radar measurements using of a cross-correlation advection field, a weighted linear combination of the corrected reflectivity measurements gathered at various elevation angles, a production of a 5-min rain accumulation using the advection field to mitigate undersampling effects, and finally a basic adjustment of those 5-min rain accumulation by computing every hour an adjustment factor from rain gauge and radar data to overcome problems linked to radar calibration.…”

Propagation of uncertainty from observing systems and NWP into hydrological models: COST‐731 Working Group 2

“…Operationally the QI is calculated for surface precipitation estimates since this is the radar-based data most often required by hydrologists (Szturc et al, 2009). QI is calculated in real-time mode in following COST-731 member countries: Germany (Friedrich et al, 2006), Poland (Szturc et al, 2008a), France (Tabary et al, 2007) and in the Emilia-Romagna region (Italy). Further ideas on assessing the quality of weather radar information for hydrological application are presented by Collier (2009b).…”