The atmosphere parameters derived from measurement data recorded by the GOME/ERS-2 and GOME-2/MetOp sensors appear in the table below. The operational GOME Data Processor (GDP) algorithm was used for the purpose. It derives the relevant values from radiation densities and from data supplied by the Polarization Measurement Device (PMD).
The GDP algorithm is based on an optimized DOAS (Differential Optical Absorption Spectroscopy) methodology for determining slant column densities, which are converted into vertical columns by applying the Air Mass Factor (AMF). Cloud parameters are derived using OCRA (Optimal Cloud Recognition Algorithm) and ROCINN (Retrieval of Cloud Information using Neural Networks) methodologies. The currently valid GDP-Version 4.x is part of the Universal Processor for UV/VIS Atmospheric Spectrometers (UPAS) which forms the basis for processing GOME data at DLR. UPAS represents a new generation of Level 2 multimission processor systems. It makes it possible to derive trace gas and cloud parameters both in near real time (NRT) and in an offline mode. UPAS uses as input Level 1 radiation densities and produces as output the total columns of trace gases and the cloud characteristics listed in the table.
For GOME/ERS-2 the operational processing chain at the Earth Observation Center (EOC) also includes generation of the Level 1 products (calibrated radiation densities) needed to derive the atmospheric parameter values. In the EPS ground segment for the MetOp mission—and as a member of the O3M-SAF —EOC is responsible for operational provision of the equivalent Level 2 parameters. For this purpose, the radiation densities derived at EUMETSAT from GOME-2 data are obtained via the EUMETCast receiver channel.
The entire processing chain for GOME-2 is incorporated in the DIMS multimission environment of EOC. DIMS also provides user services, quality control, archiving, and distribution functions. Quality routines determine for every orbit processed by UPAS the specified control parameters which provide information on the data being processed at the time, as well as the derivation of long-term statistics.