New insights from the Jülich Ozone Sonde Intercomparison Experiment: calibration functions traceable to one ozone reference instrument

Herman G. J. Smit, Deniz Poyraz, Roeland Van Malderen, Anne M. Thompson, David W. Tarasick, Ryan M. Stauffer, Bryan J. Johnson, and Debra E. Kollonige

Abstract

Although in principle ECC (electrochemical concentration cell) ozonesondes are absolute measuring devices, in practice they have several “artefacts” which change over the course of a flight. Most of the artefacts have been corrected in the recommendations of the Assessment of Standard Operating Procedures for Ozone Sondes (ASOPOS) report (Smit et al., 2021), giving an overall uncertainty of 5 %–10 % throughout the profile. However, the conversion of the measured cell current into the sampled ozone concentration still needs to be quantified better, using time-varying background current and more appropriate pump efficiencies. We describe an updated methodology for ECC sonde data processing that is based on the Jülich Ozone Sonde Intercomparison Experiment (JOSIE) 2009/2010 and JOSIE Southern Hemisphere Additional Ozonesondes (JOSIE-SHADOZ) 2017 test chamber data. The methodology resolves the slow and fast time responses of the ECC ozonesonde and in addition applies calibration functions to make the sonde data traceable to the JOSIE ozone reference UV photometer (OPM). The stoichiometry () factors and their uncertainties along with fast and slow reaction pathways for the different sensing solution types used in the global ozonesonde network are determined. Experimental evidence is given for treating the background current of the ECC sensor as the superposition of a constant ozone-independent component (I_B0, measured before ozone exposure in the sonde preparation protocol) and a slow time-variant ozone-dependent current determined from the initial measured ozone current using a first-order numerical convolution. The fast sensor current is refined using the time response determined in sonde preparation with a first-order deconvolution scheme. Practical procedures for initializing the numerical deconvolution and convolution schemes to determine the slow and fast ECC currents are given. Calibration functions for specific ozonesondes and sensing solution type combinations were determined by comparing JOSIE 2009/2010 and JOSIE-SHADOZ 2017 profiles with the JOSIE OPM. With fast and slow currents resolved and the new calibration functions, a full uncertainty budget is obtained. The time response correction methodology makes every ozonesonde record traceable to one standard, i.e. the OPM of JOSIE, enabling the goal of a 5 % relative uncertainty to be met throughout the global ozone network.