2021 Volume 99 Issue 4 Pages 1115-1125
This article describes humidity data correction based on an intercomparison between two manufacturers' radiosondes with the assessment using precipitable water vapor (PWV) derived from Global Navigation Satellite System (GNSS) signals. In addition, we propose a method to determine whether the same correction procedure can be applied if such intercomparison cannot be conducted.
During the intensive observation called Years of the Maritime Continent (YMC) - Boreal Summer Monsoon (BSM) study in 2018 (YMC-BSM 2018), intercomparison of radiosonde between Lockheed Martin LMS6 and Vaisala RS41-SGP was conducted at Laoag, Ilocos Norte, Philippines from late July to early August 2018. While their mean difference of relative humidity (RH) was better than 5 %, dry bias was confirmed for LMS6 only during clear sky daytime soundings based on comparing PWV with that derived from GNSS signals. To use different radiosonde data with the same research-quality, we developed a correction table for LMS6 RH data.
While a direct intercomparison between different radiosondes and independently developed observational tools such as a GNSS-receiver is ideal to evaluate data quality, it cannot always be performed. We obtained LMS6 radiosonde data at different site at Yap Island, Federated States of Micronesia from another field campaign, YMC-BSM 2020, where any intercomparison could not be conducted. To decide whether the same correction procedure obtained from YMC-BSM 2018 can be applied to those data, we assessed their similarity based on the relationship between specific humidity from surface meteorological station data that were obtained independently before launch and radiosonde specific humidity averaged over 300 m from the initial radiosonde measurement point. This method allowed us to confirm the same behavior between Laoag data in 2018 and Yap data in 2020; thus, we applied our correction method to RH data in YMC-BSM 2020.
