The global structure of upper troposphere‐lower stratosphere ozone in GEOS‐5: A multiyear assimilation of EOS Aura data

Abstract

Eight years of ozone measurements retrieved from the Ozone Monitoring Instrument and the Microwave Limb Sounder, both on the EOS Aura satellite, have been assimilated into the Goddard Earth Observing System Version 5 (GEOS-5) data assimilation system. This study evaluates this assimilated product, highlighting its potential for science. The impact of observations on the GEOS-5 system is explored by examining the spatial distribution of the observation-minus-forecast statistics. Independent data are used for product validation. The correlation of the lower stratospheric (the tropopause to 50 hPa) ozone column with ozonesondes is 0.99 and the (high) bias is 0.5%, indicating the success of the assimilation in reproducing the ozone variability in that layer. The upper tropospheric (500 hPa to the tropopause) assimilated ozone column is about 10% lower than the ozonesonde column, but the correlation is still high (0.87). The assimilation is shown to realistically capture the sharp cross-tropopause gradient in ozone mixing ratio. Occurrence of transport-driven low ozone laminae in the assimilation system is similar to that obtained from the High Resolution Dynamics Limb Sounder (HIRDLS) above the 400 K potential temperature surface, but the assimilation produces fewer laminae than seen by HIRDLS below that surface. Although the assimilation produces about 25% fewer occurrences per day during the 3 years of HIRDLS data, the interannual variability is captured correctly. This data-driven assimilated product is complementary to ozone fields generated from chemistry and transport models. Applications include study of the radiative forcing by ozone and tracer transport near the tropopause.

Tropospheric column ozone response to ENSO in GEOS-5 assimilation of OMI and MLS ozone data

“…The vertical resolution around the tropopause is about 1 km. Alongside the ozone data, a large number of in situ and space-based observations are included in the GEOS-5 analyses (Wargan et al, 2015). However, OMI and MLS ozone retrievals are the only data that directly modify the analysis ozone in this version of the DAS.…”

“…In general, the sensitivity decreases with decreasing altitude in the troposphere. Wargan et al (2015) provides more details on the OMI tropospheric sensitivity and the retrieval "efficiency factors", or averaging kernels, used in the assimilation. Wargan et al (2015) and Ziemke et al (2014) previously evaluated these ozone analyses relative to sondes and other satellite data.…”

“…Wargan et al (2015) provides more details on the OMI tropospheric sensitivity and the retrieval "efficiency factors", or averaging kernels, used in the assimilation. Wargan et al (2015) and Ziemke et al (2014) previously evaluated these ozone analyses relative to sondes and other satellite data. Their assessments show that accounting for measurement and model errors in the assimilation greatly increases the precision of the tropospheric ozone over other methods of obtaining gridded TCO fields.…”

“…The ozone analyses used in this study were produced using a version of NASA's GEOS-5 data assimilation system (DAS), ingesting data from the Ozone Monitoring Instrument (OMI) and MLS on the Earth Observing System Aura satellite (EOS Aura), as described in Wargan et al (2015). A brief description of the ozone data and assimilation system is provided in the following subsection.…”

Potential sources of a priori ozone profiles for TEMPO tropospheric ozone retrievals

“…predominantly occurring in these altitude ranges (e.g., Wargan et al, 2015;Ott et al, 2016). The work by Wargan et 194 al.…”