451
MEASUREMENT OF AERODYNAMIC ROUGHNESS USING RADAR
BACKSCATTER OVER VEGETATED SURFACES
S. D. Wall 1 , K. R. Rasmussen 2 , and R. Greeley 3
'jet Propulsion Laboratory, California Institute of Technology, Pasadena, CA 91101, USA; 2 Department of Earth
Science University of Aarhus, Arhus, Denmark; 3 Department of Geology, Arizona State University, Tempe, AZ
USA
ABSTRACT
Measurement of wind fields is important for many reasons. Wind regime data can be used to infer the amount and
type of wind-induced (aeolian) transport of sand and dust, or to establish global circulation models, for example on
other planets. Since local measurements are costly and often impossible, it is desired to infer such data from
remotely-sensed information. This paper describes a potential mechanism for remotely inferring the wind regime by
using synthetic-aperture radar data to describe the roughness of the surface, and describes a project to estimate the
practicality of using such a mechanism. An experiment is reported that extends the mechanism to vegetated sites,
where the goal is to measure potential for erosion.
KEY WORDS: Roughness, Wind, Aeolian Processes, Radar, Vegetation
INTRODUCTION
Aeolian transport of small particles depends on wind flow and is an important quantity to measure for several
economic-related reasons. The direct measurement of wind flow regime generally involves construction of wind
towers and many days of data collection, making such data extremely expensive and prohibitive in areas that cannot
be easily accessed. If some estimate of wind regime could be made from remotely-sensed data, important geological
and ecological problems on Earth and other planets could be addressed. Indeed, qualitative use has been made of
radar wind streaks on Venus to assess the direction of wind flow and to infer global wind patterns [Greeley et al.,
1992], but the addition of quantitative information would be a welcomed addition.
The effect of roughness on the wind field is parameterized in terms of a scaling length that, for a given surface,
determines the height at which the wind speed becomes zero. Since in fact the wind speed never reaches zero even
at the surface, a more practical parameter is the height at which extrapolation of the wind speed reaches zero,
generally represented by z 0 [Arya, 1982]. Microwave reflectivity is a function of the radar parameters used (e. g.,
wavelength, incidence, angle, and transmit and receive polarization) and the surface properties (surface topography
and complex dielectric constant). For modest topography and typical materials, the roughness at or near the radar
wavelength dominates [Blom et al., 1987\ Wall and VanZyl, 1989].
Since both radar and wind flow respond directly to surface roughness, it is reasonable to suspect that a fairly wellbehaved
wellbehaved quantitative relationship might exist between normalized radar backscatter coefficient, 0°, and z Q [Ulaby et
al., 1992]. Of course, the scales of topography that affect the wind are much broader, and only if the target area
contains no roughness at scales significantly greater than the radar wavelength could the relationship be expected to
hold dependably. In fact, however, estimations of z Q from wind profile data also require a large homogeneous fetch
in order to assure that measurements are taken within an equilibrium boundary layer.
The Radar and Aeolian Roughness Project (RARP) has been formed in order to investigate whether such a
relationship exists, to determine the relationship^) over a variety of surface types, and to seek a theoretical basis
from which to extend the relationship to surfaces that cannot be directly examined [Greeley et al., 1991a], We have
collected wind data using towers instrumented with anemometers in both desert and vegetated areas, and have
