Towards an along‐track validation of HOAPS precipitation using OceanRAIN optical disdrometer data over the Atlantic Ocean

Abstract

The Hamburg Ocean Atmosphere Parameters and fluxes from Satellite data (HOAPS) passive‐microwave precipitation scan product is compared with the Ocean Rainfall And Ice‐phase precipitation measurement Network (OceanRAIN) surface‐based precipitation reference dataset over the global ocean. For the first time, we apply statistical point‐to‐area adjustments and along‐track averaging to ship‐based OceanRAIN precipitation data over the Atlantic Ocean to represent collocated precipitation rates better within a HOAPS satellite pixel. The statistical adjustments reduce the HOAPS–OceanRAIN root‐mean‐square error strongly from 2.65 to 1.01 mm/hr. Overall, the point‐to‐area effect impacts HOAPS–OceanRAIN differences more strongly than the precipitation regime. Higher‐resolution satellite data indicate that these adjustments work best for the most convective‐like precipitation cases, while some rather stratiform‐like precipitation cases would need no adjustment. Excluding precipitation rates below the HOAPS sensitivity threshold of 0.3 mm/hr reduces the difference in average precipitation rates between HOAPS hits and false detections combined and OceanRAIN hits and misses combined to 2%. This precipitation‐rate difference lies below the uncertainty obtained from resampling of about 10%. Without false detections, the HOAPS precipitation rate of hits‐only exceeds that of OceanRAIN by 50%. Most of the HOAPS false detections follow from cases when precipitation occurs within the HOAPS pixel but off the ship track. Consequently, these apparent false detections lead to an overestimation of HOAPS precipitation rates compared with OceanRAIN, particularly in the inner Tropics and partly the midlatitudes, where clustered convective precipitation occurs most frequently. Misses cause underestimated HOAPS precipitation rates, mainly in mid and high latitudes. However, for a HOAPS validation, apparent false detections need to be considered in addition to hits of OceanRAIN precipitation rates, to which we successfully applied statistical point‐to‐area adjustments.