Howard University Beltsville Campus Ozonesonde Record: Data Homogenization and Uncertainties

Abstract

Ensuring the quality and consistency of multidecadal electrochemical concentration cell (ECC) ozonesonde records is essential because changes in sonde instrumentation, sensing solutions, operating procedures, and data-processing methods can introduce biases that compromise their suitability for trend analyses and model/satellite evaluation. Following best practices outlined by global ozonesonde experts through the Assessment of Standard Operating Procedures for Ozonesondes 2.0 (ASOPOS 2.0), we summarize the “homogenization” of the Howard University Beltsville Campus (HUBC; 39.054°N, 76.877°W, 53 m elevation) ozonesonde data set from 2006 to 2022. The HUBC station lies in an urbanized Mid-Atlantic region north of Washington, DC, serving as a prototype for most of the US east coast where mountains, megacities, the coastal plain, and large bodies of water yield complex transport patterns and challenges for model and satellite ozone estimations. We present an overview of the data reprocessing methods, their application to the HUBC time-series, and evaluate the homogenized HUBC data set relative to the original records. Total column ozone (TCO) from the reprocessed record, with uncertainties, is compared to TCO from satellite and a nearby Brewer spectrometer. The results of the homogenization: (a) sonde total column ozone increases by up to +3%; (b) stratospheric negative biases in the Aura MLS-sonde comparison improve from a mean of −3.2% before homogenization to −1.8%; (c) Total column ozone comparisons of both satellite-sonde and Brewer-sonde display similar improved percent differences with a decreased variance and reveal satellite sensitivity issues in capturing lower-tropospheric measurements; (d) sonde ozone uncertainties fall below 7% throughout profile, regardless of season.