Open AccessLimitations to a Geostationary Infrared Sounder due to Diffraction: The Meteosat Third Generation Infrared Sounder (MTG IRS)
Author(s) -
J. Grandell,
R. Stuhlmann
Publication year - 2007
Publication title -
journal of atmospheric and oceanic technology
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.774
H-Index - 124
eISSN - 1520-0426
pISSN - 0739-0572
DOI - 10.1175/jtech2055.1
Subject(s) - geostationary orbit , remote sensing , atmospheric infrared sounder , radiance , environmental science , depth sounding , geostationary operational environmental satellite , infrared , atmospheric sounding , satellite , meteorology , geology , physics , optics , astronomy , oceanography
Geostationary infrared sounding missions offer good temporal coverage; however, because of the large distance to the observed earth targets, the effect of diffraction is increased compared to sounding from a low earth orbit (LEO). Because of the wavelength dependence of diffraction, the spectral channels do not sample the same volume of air, as is generally assumed by retrieval algorithms for LEO infrared (IR) sounder data. This additional error introduced in the retrieval by diffraction-limited instruments is called pseudonoise throughout the paper. One such diffraction-limited geostationary system is the candidate Infrared Sounder (IRS) mission on the European Organisation for the Exploitation of Meteorological Satellites (EUMETSAT) Meteosat Third Generation (MTG) satellites with a planned need date in 2015, due to the expected lifetime of the current Meteosat Second Generation (MSG) satellites. A simplified point spread function (PSF) is applied. To represent the channels within natural conditions, measured spectra from the LEO Atmospheric Infrared Sounder (AIRS) are used as an underlying scene when integrating over the PSF. As the AIRS spatial resolution is 15 km, the basic assumption made was that the meteorological and surface features can be regarded as fractal to the extent that they can be downscaled to at least 4 km, making them useful for these investigations. The results show that the pseudonoise is highly dependent on wavelength, and highest in the window and CO2 regions (within the broader 700–1200 cm−1, or 8.3–14.3-μm, region). The worst-case pseudonoise values are approximately 1 K for these regions. A case study of the impact of the obtained pseudonoise values on temperature and water vapor retrievals shows that only with the worst-case assumptions (uncorrelated noise) is there a quantifiable impact on the results. For temperature retrievals, this ranged from 0.1 to 0.3 K in the lower and upper troposphere, respectively.