Robustness of spaceborne–ground radar reflectivity comparisons to conditioning coordinate and observational support: application to FY-3G PMR and an S-band radar network

Abstract. Spaceborne–ground radar reflectivity comparisons can be sensitive to how reflectivity level is represented, how comparable observations are selected, and how uneven observational contributions are aggregated. Using warm-season observations from 2024 and 2025, we volume-matched Fengyun-3G (FY-3G) Precipitation Measurement Radar (PMR) observations with those from five operational S-band radars in Liaoning Province, China, yielding more than 1.6 million matched reflectivity pairs from 146 independent FY-3G overpasses. The overall PMR–S-band reflectivity difference varied with radar composition, aggregation scheme, and warm season. Using exactly the same matched pairs, the reflectivity-dependent slope changed from −0.532 dB dBZ⁻¹ under conventional ground-radar conditioning to −0.101 dB dBZ⁻¹ under the paired-mean coordinate, and the weaker negative dependence was reproduced across observational subsets and radar sites. Stratiform and convective precipitation were then compared within the same radar–overpass task and paired-mean reflectivity interval. The task-balanced convective-minus-stratiform contrast was 4.06 dB and decreased to 2.59 dB when both reflectivities were restricted to at least 15 dBZ, while remaining predominantly positive. These results show that conditioning-coordinate choice, common observational support, and aggregation level materially influence the magnitude and interpretation of spaceborne–ground radar reflectivity comparisons.