Performance Evaluation of Polarimetric Radio Occultations Measurements in Detecting Precipitation.
Abstract: For the first time, ROHP-PAZ mission is using polarimetric radio occultations mea- surements to estimate precipitation. This paper evaluates the performance of the averaged vertical profiles ∆Φ in detecting precipitation above a certain threshold. Using IMERG precipitation prod- ucts as target, I developed an algorithm that computes the F1 Score for different averages and found that the optimal range for computing this average starts at ho <∼ 1 km and reaches hf ∼ 5 − 10 km. These averaged profiles achieve F1 Scores ranging from 0.4 to 0.6 depending on how imbalanced the data is. Moreover, the results shows how the optimal range shifts to higher altitudes when studying tropical regions due to the higher levels of noise registered near the surface at low latitudes and the vertical structure of convective precipitation which is predominant in these regions.
The PAZ Python module was developed by Ignacio Cordova Pou during 2022 for his final thesis at the Universitat de Barcelona. The module is used for the paper: "Performance Evaluation of Polarimetric Radio Occultations Measurements in Detecting Precipitation" and contains funtions to work with the data gathered in the ROHP-PAZ mission. To acces the data please contact ICE-CSIC at https://www.ice.csic.es/ .