The Jornada is also a NASA Earth Observing System Land Validation Core Site and is one of three study sites in a NASA EOS-funded research project to map carbon pools using Multi-Angle EOS data, as well as a CHRIS/Proba Core Site.
Q: Why not just use Landsat? A: Landsat can't see either big shrubs on bright desert soils or smaller shrubs encroaching into areas with more dense understories. The TM pixel is bright/dark primarily because of the background (bare sand, grassy understory, or something in between), not because of the shrubs. These images are centered on the West Well in the Jornada Experimental Range. The same is true for all medium resolution imagers: (SPOT, China-Brazil, IRS, CHRIS/Proba...). That is why we are trying to use multiangle imaging that depends on the effects of canopy structure to access woody plant abundance.
Below are scans of Kodak Advantix index prints for various types of grassland and shrubland
found at the Jornada Experimental Range. The solar zenith angles (from normal to the target
surface to the sun) were rather high and in the range 49 - 63 degrees. All acquired on 30th
May 1999 from the top of a Chevy Blazer 4WD at an incidence angle sufficient to just
include a small part of the horizon (view zeniths from about 20 to 80 degrees - the ends of
shadow of the photographer's head are about 10 to 20 feet away from the Chevy (16ft at the
mesquite shrubland site). Note that the shots taken on the antisolar (backscattering)
direction are considerably brighter than those taken in the solar (forward-scattering)
direction. This is mainly a function of geometrical effects (facet illumination and
shadow-hiding). Volume scattering effects are small in relative magnitude. The impact of
these phenomena on the radiance reaching an optical sensor scanning off-nadir and at large
angles is clearly important. Directional effects are less pronounced at lower solar zenith
angles - but are still important. The phenomena are determined by the Bidirectional Reflectance Distribution Function (BRDF) SO WHAT?
Appreciating the significance of the BRDF is one thing - but now what? Since the degree of
scene darkening is dependent mainly on the height and spacing of the shrubs (and shrub size
and cover), multiangular reflectance data should provide information on these parameters.
Prerequisites are adequate sensor calibration and correction for scattering and absorption
of light by the atmosphere (which has its own BRDF), adequate angular sampling...and a
decent BRDF model. Validation data can be provided via high resolution aerial photography.
Even with imperfect
and sparse multiangular datasets we are able to perform the two major applications of this
technology : correcting for BRDF effects in off-nadir data to improve data consistency and
deriving information on the surface which is not available in the spatial, spectral or
temporal domains.
One important application of BRDF modeling is community type and cover type mapping: the directional signal adds to the information available in the spectral domain to improve the ability to isolate the classes. This has been recognized or some time; however it is only recently that the degree of the improvement which might be expected has been realized. Recent research over semi-arid grasslands and shrublands has shown that it is possible to discriminate cover types which are quite similar spectrally (and temporally), even with moderate resolution visible and near-infrared reflectance estimates from satellites. See: Chopping, M.J., Rango, A., and Ritchie, J.C. (2002), Improved semi-arid community type differentiation with the NOAA AVHRR via exploitation of the directional signal, accepted 01/29/02 for publication in IEEE Transactions in Geoscience and Remote Sensing. See here (Inner Mongolia) and here (New Mexico).
1/2: Creosote shrubland / bare soil (7:20am; SZA=63.1
degrees)
3/4: Creosote shrubland / low grasses (7:30am SZA=61.5
degrees)
5/6: Mesquite shrubland (7:40am
SZA=58.2 degrees)
7/8: Grassland (JORNEX site,
8:05am SZA=54.9 degrees)
9/10: Transition (JORNEX site, 8:15am SZA=51.7 degrees)
11/12: Mesquite duneland from tower (JORNEX site, 8:30am
SZA=48.5 degrees)
Left : Backscattering Right : Forward-scattering
