An improved European Multi-Sensor Total Ozone Climate Data Record as Part of the ESA Climate Change Initiative
Lerot, Christophe1; Van Roozendael, Michel1; Spurr, Robert2; Loyola, Diego3; Coldewey-Egbers, Melanie3; Koukouli, MariLiza4; Balis, Dimitris4; Lambert, Jean-Christopher1; Granville, Josť1; Goutail, Florence5; Pommereau, Jean-Pierre5; Zehner, Claus6
1BIRA-IASB, BELGIUM; 2RT Solutions, Inc., UNITED STATES; 3DLR/IMF, GERMANY; 4AUTH, GREECE; 5LATMOS, FRANCE; 6ESA, ITALY
The ESA Ozone Climate Change Initiative project (Ozone_cci) aims to produce and characterize a number of high quality ozone data products generated from multiple satellite sensors. Within this framework, we have recently released a suite of improved level-2 and level-3 total ozone data sets based on the European sensors GOME/ERS-2, SCIAMACHY/Envisat and GOME-2/MetOp-A covering more than 17 years of data.
The level-2 data records were generated with the retrieval algorithm GODFIT (GOME-type Direct FITting) jointly developed by the BIRA-IASB/RT-Solutions/DLR-IMF consortium. Total columns are derived from a non-linear least squares adjustment of reflectances simulated with radiative transfer tools from the LIDORT suite, to the measured spectra in the Huggins bands (325-335 nm). In particular, the inter-sensor consistency has been greatly improved by applying a specific soft-calibration scheme on the individual level-1 reflectances, in addition to the homogenization of all input auxiliary data bases.
The level-2 data records are then combined to create a consistent total ozone ECV consisting of a time series of homogenized 1°x1° monthly mean total ozone columns. The merging procedure corrects for subsisting inter-satellite biases and temporal drifts.
We describe here the main features of these new ozone data sets, focusing on algorithmic improvements and their impact on the final product. In particular, we show that the soft-calibration scheme implemented in the level-1 to -2 chain helps to reduce significantly the correction factors needed for the merging procedure that generates the ECV. The quality of the various reprocessed total ozone data sets is assessed using independent total ozone measurements both from space and ground. Also, some comparisons with the recent 8-years MACC reanalysis are presented, with the accent on both the inter-sensor and the model-observation consistencies.