GRAS SAF -

GRAS SAF ROPIC Project The GRAS SAF is in 2006 conducting the Radio Occultation Processing Intercomparison Campaign (ROPIC). The project is carried out as a Visiting Scientist project by Armin Löscher at DMI and the purpose is to validate the processing chains from various RO centers worldwide. The idea for this investigation of the performance of different RO processing algorithms and implementations originated from discussion between Axel von Engeln, Gottfried Kirchengast, Chris Marquardt, and the GRAS SAF Steering Group. The project will also be coordinated with the ROSE (Radio Occultation Sensor Evaluation) campaign, carried out by GFZ and its partners.

The first announcement related to ROPIC can be read here.

Motivation

Now there are worldwide 7 centers processing RO data, starting from different levels, to bending angles and refractivities, some at an operational level.

DMI (Danish Meteorological Institute)
EUMETSAT (European Organisation for the Exploitation of Meteorological Satellites)
GFZ (Geo Forschungs Zentrum Potsdam)
JPL (Jet Propulsion Laboratory)
UCAR (University Cooperation of Atmospheric Research)
UK Met Office
Wegener Center, University of Graz
DWD (Deutscher Wetter Dienst)

Most of the processing centers are also generating further processed products like temperature and humidity profiles. The processing chain is composed of a number of algorithms which differ between the centers; even if the same underlying algorithm is used the specific implementations will be different, resulting in slightly different results. For that reason the idea to undertake a comprehensive processing center intercomparison was born.

Objectives

The intention is not the assessment of the absolute accuracy but to reveal possible inconsistencies between the results of different centers and the identification of benefits from using certain algorithms and methods. Most likely the optimal processing method is complex (e.g. it may vary with altitude range and spatial location of the observation).
The aim of this project is thus that we all gain a better understanding and insight in the behavior of our processing chains and to come up for discussions to optimize our procedures. Supplied data will be handled confidential; the results are foreseen to be published after giving the centers a chance for reevaluation.

Brief outline

A detailed project plan will be presented later this year (aimed for begin of March), the rough outline will be:

Compilation of a common dataset (CHAMP based) covering one month of two seasons (thus roughly two times 5000 profiles) to be used by all participating RO centres.
The data set should be as complete as possible (starting from e.g. raw data, containing several intermediate processing steps).
Since not all centres are starting their processing from the raw data, we will also provide the excess phases and orbits (Level 1A, in our terms Excess Phase & Orbits) as a common data set for all participants (provider of the Raw/Level 1A data is yet to determined).
The primary focus will be on bending angles and refractivities, further processed products as temperature and humidity will be a plus. Nevertheless the data set should be thorough as mentioned before.
The centres are asked to provide collocated reference profiles derived from ECMWF data (Operational Analysis).
Investigation of the intermediate processing steps is an option for the future.
Usage of a dedicated web-page for information and documentation; ftp access for data handling.
First evaluation and processing of the delivered data, results are available only internal.
Centres are getting a chance for revaluation and adjustments.
Final evaluation and processing of delivered data.
Publication of the results (report & journal).

Preliminary time schedule

The preparatory phase is envisaged to be terminated first week of March, meaning, that the Level 1A data and necessary documentation will be available from our web server. The centers are given, starting from the availability of data, about 2-3 months to do the processing. About 6 months are scheduled for evaluation of the comparison, including possible feedback from the centers (centers are given roughly 6 weeks for reevaluation of data and submission of final data sets). The whole project has to be finished by end of 2006.
The centers are asked to process the selected profiles from Level 1A (provided by us) to bending angles nd refractivities (the main parameters analyzed within this study; results of intermediate processing steps should be stored for possible further studies). The refractivity/bending angle data should be processed onward to temperature and humidity profiles (with collocated ECMWF profiles) for further investigations.

The preliminary schedule and project outline are subject to adjustments.