Roeland van Oss

Regridding of remote soundings : Formulation and application to ozone profile comparison

[1] One of the difficulties arising when intercomparing independent measurements of atmospheric constituent profiles consists in homogenizing their respective profile coordinates in order to allow quantitative comparison results. Special care should be devoted in particular to the homogenization of remote sensor measurements, those being indeed intricately bound to their respective numerical grids through discretization rules implied for the evaluation of the retrieval algorithms. Recently, a method of intercomparing remote sounders while allowing for different observational characteristics was proposed by Rodgers and Connor (2003). However, at the time of publication, application of this method was restricted to comparisons of identical state vectors. We propose to relax this condition by the use of linear transformation functions to homogenize the products of independent retrievals. We combine this technique with the Rodgers and Connor procedure to compare independent ozone profile measurements by the Global Ozone Monitoring Experiment (GOME) and a ground-based microwave radiometer (MW). Verification of the achieved results is obtained by considering a second series of MW retrievals, evaluated directly on the GOME numerical grid.

Wavelength calibration of spectra measured by the Global Ozone Monitoring Experiment by use of a high-resolution reference spectrum

Earthshine spectra measured by the nadir-viewing Global Ozone Monitoring Experiment (GOME) spectrometer aboard the ERS-2 satellite in the range 240 790 nm are widely in use for the retrieval of concentrations and vertical profiles of atmospheric trace gases. For the near-real time delivery of ozone columns and profiles at KNMI, a tailormade wavelength calibration method was developed. The method uses a high-resolution (0.01 nm) solar spectrum as reference spectrum and applies both a shift and a squeeze to the wavelengths in selected windows to find the optimal wavelength grid per window. This provides a calibration accuracy of 0.002 nm below and 0.001 nm above 290 nm. The new wavelength calibration method can be used on any wavelength window, for example to improve the calibration of spectra from the GOME Data Processor (GDP). A software package called GomeCal which performs this re-calibration, along with an improved polarisation and radiometric correction, has been made and is released via the WWW. The method can be used for any high-resolution (ir)radiance spectrometer, such as the satellite instruments SCIAMACHY, OMI and GOME-2.

Ozone profile retrieval from GOME data

GOME is the first satellite instrument with the possibility to retrieve height-resolved ozone densities in both stratosphere and troposphere. The high accuracy and spectral resolution of the GOME spectrometer in the range of 240-790 nm combined with sophisticated retrieval algorithms enables the derivation of accurate ozone profiles. This paper discusses in detail the retrieval procedure of ozone profiles from the GOME observations. The resulting profiles and their calculated errors are discussed and compared to local ozone profiles form ozone sonde measurements.

Spectrograph dedicated to measuring tropospheric trace gas constituents from space

Several organizations in the Netherlands are cooperating to develop user requirements and instrument concepts in the line of SCIAMACHY and OMI but with an increased focus on measuring tropospheric constituents from space. The concepts use passive spectroscopy in dedicated wavelength sections in the range of 300 to 2400 nm and wide angle, non-scanning, swath viewing. To be able to penetrate into the troposphere small ground pixels are used to obtain a fair fraction of cloud-free pixels and to allow precise detection of the sources of polluting gases. The trace gas products aimed for are O3, NO2, HCHO, H2O, SO2, Aerosol (optical depth, type and absorption index), CO and CH4, covering science issues on air quality and climate. The main challenge in the instrument design is to obtain a good signal-to-noise for cloud free pixels and for low ground albedo and light levels. Also the retrieval of separated tropospheric and stratospheric column amounts from a nadir looking instrument is challenging. The paper discusses the user requirements and compares alternative measurement strategies. It explains the selection of passive UV-Visible-NIR spectroscopy and comes with an instrument concept which provides the current best realisation of the user requirements.