×

You are using an outdated browser that does not fully support the intranda viewer.
As a result, some pages may not be displayed correctly.

We recommend you use one of the following browsers:

Full text

Title
Mesures physiques et signatures en télédétection

1227
Specific Comment: Indices and Compositing Techniques
Directional reflectance effects can be a potential source of information on surface parameters, which may be
retrieved by inversion (see comments on models and inversion). They may also be a source of noise when
one wishes to monitor in time the vegetation with wide field-of-view sensors such as AVHRR, for which
the view direction changes significantly from day to day, or when one needs to compare the reflectance
measurements of various sensors with differing geometric configurations. New indices have been proposed
to reduce or eliminate this as well as other sources of noise. For similar reasons new compositing
techniques have been proposed, including techniques based on the use of BRDF models which provide time
profiles of reflectance corrected for directional effects. The wealth of indices and compositing techniques
which have been proposed these last years is at present confusing. There is a need to establish a protocol of
validation and intercomparison of these various methods.
RECOMMENDATIONS.
1. We strongly recommend continued research into the application potential of the BRDF and polarization
properties of ground surface and atmospheric targets. The research objectives should address the
information needs of potential users.
2. Three dimensional models, which account for the statistical behavior of remotely sensed data, should be
extended and inverted in order to support analysis of data potentially covering rolling terrain such that
pixels represent heterogeneous mixtures of surface cover types and project ground footprints with sizes
between 10m and 6 km — the ground pixel sizes of planned future sensors.
3. Available reflectance models should be validated further by means of multi-dimensional (directional,
spectral, temporal) field data. Existing models should also be intercompared in more depth to evaluate
their performance and limitations.
4. Existing methods for model inversion should be validated in more depth in order to quantify the
practical limitations and the expected accuracy of the parameters retrieved. New approaches should be
developed based upon a priori knowledge of plant canopy development and spectral BRDF properties.
5. There is a need to establish a protocol of validation and intercomparison of the indices and
compositing techniques which have been proposed during these last years.