|
|
Evidence of pre-launch characterization problem of Meteosat-7 visible spectral response curve This publication appears in: Remote Sensing Letters Authors: I. Decoster, N. Clerbaux, E. Baudrez, S. Dewitte, A. Ipe, S. Nevens, A. Velazquez Blazquez, J. Cornelis and Y. Govaerts Volume: 4 Pages: 108-117 Publication Date: Jul. 2013
Abstract: Since the early 1980s, the geostationary Meteosat instruments have been observing
the earth at a high temporal and spatial resolution. The data record contains more
than 30 years of observations and has become very useful for climate monitoring.
With a compilation of 10 satellites, it is crucial to reach the highest possible consistency
between the instruments. However, the quality of the level 1 data records
is still far from what is needed to generate homogeneous climate data records, due
to both temporal ageing as well as discontinuities between successive instruments.
In this letter, a method is proposed to assess the pre-launch Meteosatlj visible
spectral response curve and a model of its ageing. The results not only validate
the ageing model, but also provide evidence that the official pre-launch spectral
response curve of Meteosatlj poorly represents the sensitivity of the instrument in
the blue part of the visible spectrum. Until theMeteosatlj visible spectral response
curve has been reassessed, it is suggested to replace the Meteosatlj curve with the
curve of the high resolution visible channel ofMeteosatNJ. Quantitatively, the error
introduced when unfiltering the data (i.e. conversion from filtered to unfiltered radiance),
due to the uncertainty on the Meteosatlj spectral response, is estimated at
4.5% using the Meteosatlj spectral response curve, but could be reduced to 2.1%
by following the recommendations in this letter.
|
|
