How QPE algorithms can modify the symmetries of radar data

Scaling symmetries of rainfall have been too often seen as theoretical concepts, whereas we demonstrate their practical use for radar products validation. This is increasingly important with the emergence of new radars technologies, such polarimetric X-band radars, which enable local authorities and private companies to run their own radars, especially in urban areas where stakes are high. Meanwhile researchers and practitioners were mostly focusing on the development of new Quantitative Precipitation Estimation (QPE) algorithms, aiming a common goal of going as far as possible on the correction of radar data, However, such QPE algorithms may lead to unexpected modifications of the scaling symmetries of the original raw data. This study will present a recent research done in Paris area involving the comparison of two radar products using both the same raw C band radar data but relying on different QPE algorithms. Results show how the use of some processing algorithms can modify the scaling symmetries of the final product, leading thus to significant changes of the morphology of radar data. The comparison was performed using 5 rainfall events and with the help of Universal Multifractals applied in both spatial and temporal analysis.