On the quality of the ERA-Interim ozone reanalyses. Part II Comparisons with satellite data

Title
On the quality of the ERA-Interim ozone reanalyses. Part II Comparisons with satellite data
Report
Date Published
05/2010
Series/Collection
ERA Report Series
Document Number
3
Author
Rossana Dragani
Event Series/Collection
ERA Report
Abstract This is the second of two companion papers presenting an assessment of the quality of the ERA-Interim ozone reanalyses by comparisons with independent observations, during the period January 1989 to December 2008. Ozone profiles from SAGE, HALOE, UARS and Aura MLS, POAM II and III were used to validate the three-dimensional ERA-Interim ozone analyses. Total column ozone (TCO) from OMI and a global mean TCO reference were used to assess the quality of the TCO product. The ERA-Interim TCO is typically within ±5DU (about ±2%) from the TCO reference, while showing up to 2% lower values than OMI TCO between 50◦S-50◦N. Comparisons with SAGE, HALOE and MLS ozone profiles showed some degrees of consistency in the results. The ERA-Interim analyses at and above 10hPa were typically within ±10% from SAGE and HALOE, and within ±5% from MLS. Around 30hPa, positive analysis departures up to +25%and up to +35%were found before January 1996 in the cases of SAGE and of HALOE andMLS, respectively. With the active assimilation of GOME ozone profiles (January 1996 through December 2002), the level of agreement typically improved leading to residuals mostly within ±10%. Positive departures up to 20% were found in the lower stratosphere (around 70hPa), except in the tropics during the GOME ozone assimilation period when these residuals were within -5 and +10%. The comparisons at high latitudes with POAM data showed discrepancies larger than 50% at and above 10hPa, and typically within ±20% below. The analysis departures showed trends typically within ±0.5%/year. When statistically significant, these changes were normally associated to an improved agreement with time between the ERA-Interim ozone analyses and the independent data. The comparisons also showed higher quality of the ERA-Interim ozone analyses compared with the ERA-40 ones, particularly due to the assimilation of GOME ozone profiles.