2 Physical Principles and Technical Aspects of Remote Sensing
35
eters, optimized for measuring backscatter from water drops and ice particles
in precipitating clouds with vertical resolution of several hundreds of meters
and horizontal resolution of a few kilometers. The first spaceborne rain radar
is flown on the Tropical Rainfall Measuring Mission (TRMM).
Space borne imaging radars (Elachi, 1987) provide surface images with
high spatial resolution (between about 10 m and 100 m) over a swath of
about 100 km to a few hundreds km width. The across track resolution is determined by the pulse length (Eq. 2.27), high resolution is achieved by means
of pulse modulation. The ground resolution across track, r g, varies with the
incidence angle: rg = rr/sinOi (Fig. 2.15). Synthetic aperture techniques are
applied to obtain high resolution along track. The Doppler shift of the return
signal, which varies along track within the illuminated footprint, is utilized
to synthesize an antenna of 1 km length or more. Radar echos from many
hundred pulses are added coherently to calculate the return for one pixel.
Theoretically, the along track resolution, r a, depends only on the antenna
length (ra = L/2) and thus is independent of the distance between antenna
and surface. However, there are practical limits resulting from the power required to detect a return signal. Typical antenna lengths of spaceborne SARs
are about 10 m. Swath widths with conventional antenna design are ~ 100
km. Distributed array technology (e.g. on Radarsat) applies electronic beam
steering to cover a wider swath with reduced resolution or to select a narrow
sub-swath with full resolution.
The radar echo from an illuminated surface area or volume element is
the coherent addition of the returns from the individual scatterers. Therefore
the resulting return signal of a distributed target that consists of many point
scatterers fluctuates randomly in successive observations as the radar moves
along. The statistics of the backscattering power, P, which is proportional to
(70, is described by an exponential probability density function (Ulaby et al.,
1982). The standard deviation of the power is
(7p = P/VN
(2.28)
where N is the number of independent looks and P is the mean power.
N = 1 is valid for pixels in full spatial resolution. In order to measure (70 of a
distributed target accurately, averaging or low pass filtering over many looks
is necessary.
Whereas single channel SARs are operating on the European Remote
Sensing Satellite (ERS) and on Radarsat, the experimental three-frequency
polarimetric Spaceborne Imaging Radar-C (SIR-C)/X-SAR has been operating during two Shuttle missions in 1994. Polarimetric SARs are also in operation on aircrafts. Radar polarimetry offers improvements for quantitative
inversion of target properties and for classification. For example, polarimetric measurements enable the separation of surface roughness and moisture
effects and therefore are of interest for soil moisture monitoring (Dubois et
al.,1995).
35
eters, optimized for measuring backscatter from water drops and ice particles
in precipitating clouds with vertical resolution of several hundreds of meters
and horizontal resolution of a few kilometers. The first spaceborne rain radar
is flown on the Tropical Rainfall Measuring Mission (TRMM).
Space borne imaging radars (Elachi, 1987) provide surface images with
high spatial resolution (between about 10 m and 100 m) over a swath of
about 100 km to a few hundreds km width. The across track resolution is determined by the pulse length (Eq. 2.27), high resolution is achieved by means
of pulse modulation. The ground resolution across track, r g, varies with the
incidence angle: rg = rr/sinOi (Fig. 2.15). Synthetic aperture techniques are
applied to obtain high resolution along track. The Doppler shift of the return
signal, which varies along track within the illuminated footprint, is utilized
to synthesize an antenna of 1 km length or more. Radar echos from many
hundred pulses are added coherently to calculate the return for one pixel.
Theoretically, the along track resolution, r a, depends only on the antenna
length (ra = L/2) and thus is independent of the distance between antenna
and surface. However, there are practical limits resulting from the power required to detect a return signal. Typical antenna lengths of spaceborne SARs
are about 10 m. Swath widths with conventional antenna design are ~ 100
km. Distributed array technology (e.g. on Radarsat) applies electronic beam
steering to cover a wider swath with reduced resolution or to select a narrow
sub-swath with full resolution.
The radar echo from an illuminated surface area or volume element is
the coherent addition of the returns from the individual scatterers. Therefore
the resulting return signal of a distributed target that consists of many point
scatterers fluctuates randomly in successive observations as the radar moves
along. The statistics of the backscattering power, P, which is proportional to
(70, is described by an exponential probability density function (Ulaby et al.,
1982). The standard deviation of the power is
(7p = P/VN
(2.28)
where N is the number of independent looks and P is the mean power.
N = 1 is valid for pixels in full spatial resolution. In order to measure (70 of a
distributed target accurately, averaging or low pass filtering over many looks
is necessary.
Whereas single channel SARs are operating on the European Remote
Sensing Satellite (ERS) and on Radarsat, the experimental three-frequency
polarimetric Spaceborne Imaging Radar-C (SIR-C)/X-SAR has been operating during two Shuttle missions in 1994. Polarimetric SARs are also in operation on aircrafts. Radar polarimetry offers improvements for quantitative
inversion of target properties and for classification. For example, polarimetric measurements enable the separation of surface roughness and moisture
effects and therefore are of interest for soil moisture monitoring (Dubois et
al.,1995).