Environment contribution to the atmospheric correction for Landsat-MSS images

The signal returning from the Earth's surface to the satellite is modified by the atmospheric effect, which has two components. The first one is solar radiation which, due to backscattering, is deviated in the direction of the sensor without reaching the Earth's surface. The second compone...

Ausführliche Beschreibung

Gespeichert in:
Bibliographische Detailangaben
Veröffentlicht in:International journal of remote sensing 1995-09, Vol.16 (14), p.2515-2537
Hauptverfasser: MILOVICH, J. A., FRULLA, L. A., GAGUARDINI, D. A.
Format: Artikel
Sprache:eng
Schlagworte:
Online-Zugang:Volltext
Tags: Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
Beschreibung
Zusammenfassung:The signal returning from the Earth's surface to the satellite is modified by the atmospheric effect, which has two components. The first one is solar radiation which, due to backscattering, is deviated in the direction of the sensor without reaching the Earth's surface. The second component is produced by the energy reflected in areas close to the pixel observed which, owing to collisions with atmospheric constituents, is deviated from its path in the sensor direction. This is called the adjacency effect and this paper presents a numerical method to estimate this effect under the assumption of a heterogeneous flat Lambertian surface. From this estimation it is possible to apply the atmospheric correction for the calculation of reflectance images based on data obtained by the optical channels of high resolution satellite systems such as Landsat-MSS, Landsat-TM and SPOT/H RV. In particular, in this paper the method is applied to Landsat-5 MSS images over urban regions. However, its application to any of the sensors mentioned is easily implemented considering the changes in spectral response and pixel size. Differences obtained in the results for reflectance at Earth's surface in winter and summer images were in the order of 10 −3 for bands 1 and 2, 10 −2 for band 3, and 10 −1 for band 4.
ISSN:0143-1161
1366-5901
DOI:10.1080/01431169508954574