565 
4 - THE SPECIAL IMPORTANCE AND NEEDS FOR DIRECTIONAL MEASUREMENTS TODAY 
Many remote sensing scientists today are focusing on measurements and/or modeling of the angular reflectance 
properties of Earth surface covers, although not necessarily at the expense of an analysis of the spectral, spatial 
or temporal reflectance characteristics (each of which yields different information about the surface covers). 
The angular studies emphasis is occurring for several reason. First, of these four primary aspects of land 
surface reflectance, this is the still the least intensively studied, and it is certainly the least understood. Second, 
it is becoming widely recognized that the angular reflectance properties of Earth targets and, in particular, the 
BRDF of land surface cover types, must be studied to develop algorithms to provide more accurate estimates 
of spectral hemispherical reflectance and albedo, which is a crucial input to Global Climate Models. Third, 
the maximum potential of current satellites (e.g., SPOT, AVHRR) that acquire off-nadir viewing data cannot 
be realized without a good understanding of the angular reflectance properties of the Earth’s surfaces. And 
fourth, future EOS satellite instruments (e.g., MISR) have been developed specifically to take advantage of, 
or to help overcome, the Earth’s surface and atmosphere angular reflectance dependencies. 
Empirical and theoretical remote sensing investigations have typically been conducted independently 
or have reached the point of specialization such that researchers either collect data or develop models. In 
general, there has been rather poor communication between these groups and, consequently, model-required 
parameter inputs (for driving the models) or outputs (for validating the models) are not measured by the 
experimentalists. Missing data are usually in the form of either incomplete spectral bidirectional reflectances 
(e.g., missing background material reflectances or inadequate angular measurement series) or associated 
biophysical variable measurements (e.g., no leaf optical properties, LAI, or plant canopy architecture measures), 
or both. 
There have been some attempts to bring the experimentalists and modelers together to try to develop 
a better understanding and to develop collaborations. These efforts have been in the form of both workshops 
(Gauthier, et al., 1991) and joint projects involving those who make measurements and those who make models 
and want to use the measurements of others. Based on my understanding of the problems and potentials through 
participation in various workshops and other discussions as well as my own field measurement experience and 
association with modeling and the current needs, the following are my suggested priority areas for future 
measurements of directional radiances: 
1) Multi-temporal directional spectral measurements, both inside and above the vegetation canopy, 
over a wide range of view and illumination angles and under a variety of sky conditions, with a complete 
characterization of the angular properties of the downwelling hemispherical irradiance. The multi-temporal, 
diurnal observations of surfaces would help define the factors controlling architectural and phenological changes 
that produce the BRDF dynamics and would produce the complete BRDF data sets necessary of validating the 
plant canopy models. Measurements under very clear and very diffuse sky conditions could provide a unique 
understanding of the role of shadowing in determining the anisotropic response. 
2) Extensive vegetation architecture measurements. These measurements are essential to appropriate 
validation of today’s plant canopy models, but they must be coupled with the more customary ancillary 
measurements obtained for the vegetation, including the leaf optical properties, plant structural component 
reflectances (e.g., bark reflectance) and background reflectances, as well as biophysical measurements of LAI 
and biomass. 
3) Multiple spatial scales (scale length) and the effects ofIFOV. The capability to successfully scale 
up or down between ground level and aircraft or satellite measurements needs to be examined. There are 
obvious limitations in the number of ground measurement sites that can be sampled, particularly in regional and 
larger-area studies of surface level biological and ecological parameters. Therefore, a closely related issue that 
merits attention is the development of spatio-physical models that allow for the accurate extrapolation of specific 
site data over the complex landscape mosaic. 
4) High spatial resolution measurements of the hotspot region. The links between the vegetation type 
and condition and the vegetation canopy architectural parameters as related to the hot spot shape and amplitude 
should be established. An IFOV of less than 5° will likely be required to characterize these hot spot properties. 
A variety of land surface cover types should be examined. The hot spot occurring at various solar zenith angles 
should be measured, as it is probable that the hot spot characteristics will change with solar zenith angle (and 
possibly its azimuth in some cover types) for the same land surface cover type and condition. 
A better understanding of the radiant interactions with the Earth surfaces is needed to develop better 
sensors and better algorithms for Earth observing satellites. Directional radiance measurements obtained from 
ground, aircraft and satellite instruments will be necessary to develop the needed improvements in our ability 
to remotely measure an monitor changes in important landscape parameters. Although numerous directional 
radiance/reflectance data sets of various surfaces exist, currently nearly all of them lack crucial observations 
of model inputs and outputs, such that their use in model validation work is limited. Better communication and 
collaboration between experimentalists and modelers will be required to ensure that the maximum potential can 
be realized from the limited resources being made available for Earth biosphere remote sensing.