Rain gauges and X-band radar hourly comparison under complex orographic conditions in Reunion Island during the passage of the cyclone Batsirai

Abstract. Weather radar observations and quantitative precipitation estimation (QPE) are in the early stages of development in the South-West Indian Ocean (SWIO) region, which is prone to heavy rainfall, particularly during the passage of tropical cyclones. Given the topography of SWIO islands, orography plays an important role in the spatial distribution of precipitation. The ESPOIRS project was designed to investigate such dynamics in Reunion Island, Seychelles, and Madagascar using a mobile X-band radar. Reunion Island served as a testbed to evaluate X-band radar retrieved QPE using specific comparisons between several radar approaches and available rain gauges. This is the first study to use an X-band dual-polarization radar in the SWIO region. Our research focuses on the intense tropical cyclone event Batsirai in Reunion Island and shows the effectiveness of dual-polarization radar when compared to single-polarization radar in mitigating reflectivity attenuation. Both the Hitschfeld and Bordan and the philinear algorithms were employed and evaluated for this purpose. As our study encountered challenges related to noisy and low-resolution differential phase (𝜙_𝑑𝑝) data, we detailed the pre-processing steps used to extract reliable 𝜙_𝑑𝑝 data from the observed measurements. Furthermore, we tested two precipitation estimators, R(Z) and R(kdp). We observed that the accuracy of R(Z) depends on the attenuation correction method. Additionally, using the extracted 𝜙_𝑑𝑝, we calculated an empirical model for R(kdp) for Reunion Island. This model provided better results compared to the R(Z) estimates, which can be explained by the fact that kdp is directly linked to precipitation concentration and does not require attenuation correction. Our findings highlight that the accuracy of the radar QPE is strongly influenced by local topography, which in turn governs local rainfall patterns, while the accuracy of QPE also depends on the type of precipitation.