23B.4 Using Data Quality to Improve Comparison between GPM Measurements and Ground Radars

Thursday, 31 August 2017: 4:45 PM
St. Gallen 1&2 (Swissotel Chicago)
Irene Crisologo, Univ. of Potsdam, Potsdam, Germany; and R. A. Warren, K. Muehlbauer, and M. Heistermann

Coinciding monsoon and typhoon seasons in the Philippines guarantee torrential and sustained rainfall for the most part of the year, increasing the risk for several disasters like flooding and landslides. Accurate and reliable monitoring of rainfall using weather radars is therefore a necessity for early warning systems. The low-density hydrometeorological monitoring networks in the Philippines makes calibration of the ground radars challenging, so we consider other sources for validation.

In this study, we explore the potential of using observations from the Ku-band precipitation radar (KuPR) on board the Global Precipitation Measurement (GPM) mission satellite to quantify calibration error for the Subic radar in the Philippines. The three-dimensional volume-matching algorithm of Schwaller and Morris (2011) is used to compare KuPR and Subic reflectivity observations for all satellite overpasses with precipitation during a three-year period. To preserve as much raw data as possible, pre-processing techniques such as clutter correction are skipped. Instead, the quality of data for each radar pixel is evaluated using a single index which incorporates beam altitude, pulse volume, beam-blockage fraction, and probability of detection.

Matching results show that calibration of the radar varies over time, while maintaining a general underestimation. The use of quality filters improves the correlation coefficients for most overpasses, suggesting the effectiveness of removing low quality pixels from the comparisons.

The use of quality filters in removing low quality data as an additional method to improve calibration measurements between ground radars and satellites, or any other reference data, is suggested.

- Indicates paper has been withdrawn from meeting
- Indicates an Award Winner