Contributions in Gothenburg proceedings

Looking down to Earth in the New Millennium,
Proceedings of the ERS-ENVISAT Symposium, 16-20 October 2000, Gothenburg, Sweden, ESA publication SP-461 (CD-ROM): <=== Publications page

 


 

Piters, A.J.M., Van der A, R.J., Van Geffen, J.H.G.M., Van Oss, R.F. and Valks, P.J.M.: 2000,
"Retrieving spectral reflectivities from Extracted GOME Instrument header data"

Abstract
 
The Extracted GOME Instrument header (EGOI) data, which are available near-real time and contain parts of the UV-Vis spectra, are used in the retrieval of ozone columns and profiles. This work is done in the framework of the ESA-DUP project GOFAP. The level 0-1 algorithms used are based on the GOME Data Processor (GDP) algorithms of DLR, but there are several differences. The corrections for straylight and polarisation sensitivity differ from the GDP algorithms in order to account for the limited spectral information in the EGOI data (only nine small spectral windows are available). Detailed validation studies have resulted in several improvements in the level 0 to 1 algorithms. These comprise the wavelength calibration, correction of Peltier cooler interference, correction for polarisation sensitivity and correction of the effective reflectivity degradation. This paper gives an overview of the level 0-1 processor.
 
Introduction
 
The aim of the project "GOME Ozone Fast Delivery and value-Added Products'' (GOFAP) is to develop a Fast Delivery (FD) processor retrieving total ozone columns and stratospheric profiles from the Extracted GOME Instrument header data (EGOIs), which are used at ESRIN for the instrument performance and quality assurance monitoring. The project is performed within the framework of ESA's Data User Programme (DUP).
 
The GOME FD processor constructs reflectivity spectra from the EGOI data ("level 0-to-1 processing"), and retrieves from these reflectivity spectra the ozone columns and stratospheric profiles ("level 1-to-2 processing"). The ozone columns and profiles are distributed via WWW within three hours after observation. For more details on the GOFAP project, see [1]. The profile retrieval is decribed in [2] and the ozone column retrieval in [3].
 
The algorithms used to construct reflectivities from the raw GOME data are based on the GOME Data Processor (GDP) algorithms [4,5]. However, there are several differences. The corrections for straylight and polarisation sensitivity differ from the GDP algorithms in order to account for the limited spectral information in the EGOI data (see Table 1). Furthermore, adapted algorithms have been used for wavelength calibration, correction of Peltier cooler interference, correction for polarisation sensitivity, radiometric calibration correction and correction of the effective reflectivity degradation. In this paper an overview is given of the level 0-to-1 algorithms. The main differences with the GDP algorithms are explained in some detail.
Table 1. EGOI spectral windows
WINDOW RANGE (NM) WIDTH (NM)
1
2
3
4
5
6
7
8
9
272.2 - 275.9
282.9 - 285.6
292.5 - 302.9
305.3 - 307.9
312.0 - 314.5
323.1 - 336.2
351.6 - 352.8
372.3 - 373.4
758.1 - 778.4
3.7
2.7
10.4
2.6
2.5
13.1
1.2
1.1
20.3

 
<=== Publications page  |  Top of this page

 


 

Van der A, R.J., Eskes, H.J., Van Geffen, J.H.G.M., Van Oss, R.F., Piters, A.J.M., Valks, P.J.M. and Zehner, C.: 2000,
"GOME Fast delivery and value-Added Products (GOFAP)"

Introduction
 
In the framework of the ESA-DUP project GOFAP [1] the delivery of GOME ozone products in near-real time (NRT) tousers has been accomplished. Since 1998 NRT ozone columns [2] are available through the KNMI website and since the beginning of 2000 they are accompanied by NRT stratospheric ozone profiles. Numerical weather prediction is expected to benefit from assimilation of stratospheric ozone profiles, if they are retrieved within 3 to 6 hours afterobservation. This requires a fast retrieval algorithm and a near-real time availability of the spectra. In this paper it is demonstrated that operational near-real time ozone profile retrieval for the use in weather forecast models is feasible.
 
The NRT ozone products of GOFAP are retrieved from specific segments of the full GOME spectrum contained in theEGOI files. These files are originally meant for instrument monitoring, but soon their suitability for NRT delivery of GOME products was recognised. Since the EGOI files contain raw data, a level 0-1 processor has been developed,partly based on the operational GDP processor.
 
For ozone profile retrieval the UV part (270-340 nm) of the GOME spectrum is used. The spectrum is measured with an integration time of 1.5 seconds, except for the spectrum below 283 nm, which has an exposure time of 12 seconds. Since GOME only measures spectra on the day side of each orbit, this implies that the processing time for one profile should not take longer than 24 seconds on average to keep up with the observations. A horizontal resolution of about 250 km suffices for weather forecasts (based on scales of nominal weather systems). The accuracy of the ozone profiles in the stratosphere should be better than the climatology presently used in the forecast models. Furthermore, a correct estimate of the accuracy is essential. There are no such requirements on the accuracy of the tropospheric part of the profile. To achieve a high accuracy in the stratosphere, several steps in the level 0-1 had to be improved, as shown by detailed validation studies on the quality of this level 1 data.
 
This paper will provide an overview of the accomplishments of the project, while details of the work will be given indedicated papers.

 
<=== Publications page  |  Top of this page

 


 

Van der A, R.J., Piters, A.J.M., Van Oss, R.F., Valks, P.J.M., Van Geffen, J.H.G.M., Kelder, H.K. and Zehner, C.: 2000,
"Near-real time delivery of GOME ozone profiles"

Abstract
 
A Fast Delivery processor has been developed to provide GOME ozone products within 3 hours after observation. This service meets the growing demand for ozone products for purposes like assimilation in numerical weather prediction models, UV radiation forecasts, and planning of experiments related to atmospheric chemistry research. The ozone products consist of total ozone columns, the global assimilated ozone field, global stratospheric ozone profiles, cloud fraction, cloud top pressure and the aerosol absorption index. The validation and current developments of the near-real time processor and the delivered ozone products are discussed.
 
Introduction
 
In the framework of the ESA Data User Program a Fast Delivery processor has been developed to provide near-real time (i.e. within 3 hours after observation) ozone products from the satellite instrument GOME [1] on ERS-2. By making maximum use of the existing ERS ground segment and European Internet links, the retrieval of ozone products can be performed within 3 hours after observation from the Extracted GOME Instrument Header (EGOI) Data.
 
A spectral and radiometric calibration is performed on the raw EGOI data to calculate the earthshine and sunshine spectra (i.e. the level 1 product) [2]. This level 1 product contains parts of the ultra violet spectrum applicable for the retrieval of ozone columns and ozone profiles and parts of the visible spectrum used for retrieval of cloud and aerosol information. The cloud fraction and cloud top pressure are calculated with the FRESCO algorithm [3]. The aerosol absorption index is computed to have an indication for areas with heavy aerosol pollution. This index is comparable to the aerosol absorption index of TOMS [4]. After processing all products are directly available on the web site http://www.knmi.nl/gome_fd/

 
<=== Publications page  |  Top of this page

 


 

<=== Post-doc. research at KNMI page

 
Jos van Geffen -- Home  |  Site Map  |  Contact Me

created: 29 January 2002
last modified: 18 March 2011