reflectance_De_Survey_2006.html   Stand 14-MAR-2006

Abstract


reflectance variation with sun zenith angle during day    film_20060108_europa.gif  4x 6:00  - 18.00 UTC
  

To derive proper bidirectional reflection distribution
functions for the AVHRR long time data series, a comparison
of brdf-effects between AVHRR data and MSG data is performed.
As the data of the geostationary satellite are available 
every 15 minutes, the variation of measured reflectance due 
to the effect of varying solar zenith angle can mostly be 
separated from changes in surface characteristics.
Near the horizon of the Meteosat disk, extreme conditions of 
forward and backward scattering allow to test the validity 
of used brdf functions.
It is shown, that the maximum in reflectivity occurs not at 
noon but is shifted toward hot spot conditions (i.e. when 
satellite and sun have least azimuth difference).
For selected test areas, cloudy data are eliminated by a 
homogeneity test. An 'experimental' brdf_function (Lambert 
reflection / mean value) is compared with analytical
brdf-functions.
These analytical functions are derived from fits of data 
sets, found in the literature (Walthall et.al, Suttles et 
al., Lativovic et al.).
These functions obey the Helmholtz law, i.e. the are 
symmetrical in solar illumination and satellite observation 
geometry.

 MSG: Variation of solar zenith angle fast
at rims extreme forward/backward observation conditions:

Example 1: film_20060108_kugelg.gif  03:00 - 21.00 UTC  4x

Click into image to start loop:

film_20060108_europa.gif  4x 6:00  - 18.00 UTC

For land surfaces, the maximum in reflectance is seen from the satellite not at noon, when the sun is at its highest position, but when

satellite and sun look in the same direction. This situation is known as hot spot condition and it is the more pronounced, the more the surface is structered..

This can be demonstrated in test areas at the eastern resp. western edge of the Meteosat disk.

As an example, the desert area Cyrenaike in the north east of Lybia shows an increase in reflectance from 18% to 30 % during the course of day, if it is considered as Lambert's surface and only a cos(zeta_sun) ( green curve) correction is applied.only the . A homogeneity test over 3*3 pixels is applied to detect cloudy situations (dashed lines). In the lower graph, additionally, the NDVI (Normalized differential vegetation index), the surface temperature in the 10.8 um channel and the temperature difference between 12.0 um and 10.8 um channel are plotted. The pink curve gives the empicical bidirectional reflection function, calculated as Lambert`s reflectance over the daily mean value. The blue curve gives an analytical brdf.

Applying this brdf-function results in the brdf_normalised reflectance (blue curvi in upper graph).

Extreme viewing conditions

Sun glint before South America ans Madagaskar

Example:MSG: film_20060224_kugelg.gif  03:00 GMT until 21:00 GMT

film_20060224_europa.gif  06:00 bis 17:45

Berlin Tagesgang



2 successive overpasses of AVHRR

In AVHRR areas are sometimes seen in adjacent orbits under different viewing conditions
Example Sahara desert on 24-FEB-2006

Reflectance ch vis008 in MSG and AVHRR 24-Feb-2006 of Sahra desert: 31 °North, 1.5°East

AVHRR:Orbit 19072 (Easterly)  Overpass   9:56,   zeta_sat= 62.7°, zeta_sun=50.8°. azi_sat=  94°, azi_sun=138.0°  (east), back-scattering
AVHRR:Orbit 19073 (Westerly) Overpass 11:27,  zeta_sat=-61.7°, zeta_sun=41.1°. azi_sat=287°, azi_sun=170.0°  (east), forward-scattering

The difference in  brdf-reflectance depends on surface structure










Kugel 20060202 composit IR + VIS: