Polarimetric X‐band weather radars for quantitative precipitation estimation in mountainous regions

Abstract

In order to minimize the economic and social impact of hazardous weather in the Alps, three X‐band dual‐polarization radars were deployed in the southeast of France during the RHYTMME project in 2011–2013 and more recently have been incorporated formally into the conventional radar network already operating S‐ and C‐band radars. In this mountainous region, the radars' beam shielding by the complex terrain, the very different altitudes of the radars, the significant attenuation at X‐band and the low density of rain‐gauges have made the integration of these radars into the operational service a challenging exercise. In this article, the framework used to evaluate the corrections applied to the volume data produced by the radars is presented and the residual errors related to the partial beam blockages, the attenuation by rain, the wet radome, and the melting layer are carefully studied and disentangled. The results of the analysis suggest that: (a) the beam blockage correction requires the use of a much higher resolution digital elevation model (DEM) in mountainous regions than the 250 m resolution DEM currently uses, (b) the dual‐pol reflectivity (Zh) attenuation correction used at Météo‐France performs well in the rain but the differential attenuation on Zdr is overestimated, (c) the attenuation of wet snow is underestimated and a mean error of 2 dB is induced by the melting layer in spring and autumn, (d) the wet radome, as well as producing significant attenuation on Zh, induces differential attenuation on Zdr. The attenuation in the melting layer is further investigated taking advantage of the height difference between the radar sites. A new specific attenuation coefficient for wet snow is suggested and it is shown to reduce the seasonal bias on three particular events observed in spring and autumn.

Radar Remote Sensing of Precipitation in High Mountains: Detection and Characterization of Melting Layer in the Grenoble Valley, French Alps

“…The choice of the X-band frequency is challenging in the considered context due to its sensitivity to attenuation [5]. First performance assessment of the RHyTMME radar network [6], points out: (1). the need to better understand and quantify attenuation effects in the ML, (2).…”

Combined use of volume radar observations and high-resolution numerical weather predictions to estimate precipitation at the ground: methodology and proof of concept

“…These errors are of three kinds: (i) measurement errors of the radar moments (Z, Z dr , φ dp ), (ii) conversion errors in the quantitative precipitation estimation of the precipitation from the radar moments and (iii) extrapolation errors in the determination of the precipitation falling at the ground from the estimations obtained at beam heights. In recent years, significant progresses have been made to reduce the first two types of errors by better controlling the quality of the polarimetric parameters (calibration of Z and Z dr , adaptive smoothing of φ dp , attenuation correction) (Bringi et al, 2001;Gourley et al, 2009;Yu et al, 2018) and by combining the polarimetric moments to estimate the precipitation with minimal uncertainty (Ryzhkov et al, 2005;Tabary et al, 2011;Figueras i Ventura et al, 2012). However, the final step of estimating the precipitation at the ground remains a major challenge in particular in mountainous regions where lower beams can be partially or totally blocked (Creutin et al, 1997;Smith, 1998).…”

Correcting Radar Rainfall Estimates Based on Ground Elevation Function

“…Figure 1 illustrates some of the errors that arise from the comparison between radar rainfall estimates and ground rainfall measurements using a rain gauge. Indeed, evaluation of the radar rainfall estimates accuracy based on ground rainfall only account for all of the combined error factors and does not provide information about the individual sources of error (Yu, N, et al, 2018). Several previous studies that using G/R comparison methods in their analysis, including Burcea, S, et al, (2012) Krajewski, W.F, et al, (2010), Sebastianelli, S, et al, (2013).…”

“…In the mountainous region, radar rainfall estimates accuracy is limited by partial beam blockage and non-uniform beam filling due to topographic effect ( (Yu, N, et al, 2018;Young, C.B, et al, 1999;Shakti, P.C, et al, 2012;Germann, U, et al, 2006). The X-band radar is practically useful to overcome these limitations, including miscalibration of radar hardware and attenuation.…”

“…X-band radar has several advantages compared to S and C band radars, involving higher spatial and temporal resolution, smaller antenna size, lower transmitted power for the same sensitivity, and lower costs (Orellana-alvear et al, 2019;Yu et al, 2018;Park et al, 2005). However, X-band radar is still unable to avoid the attenuation caused by heavy rainfall, where the signal received by radar is lower than the noise level (Burcea, S, et al, 2012;Hirano, K, et al, 2014;Shi, Z, et al, 2017;Yoon, S.-S. & Bae, D.-H, 2013).…”