IEEE Geoscience and Remote Sensing Magazine - June 2022 - 30

parameters enabled by SAVERS-TDS, and the validation effort
based on the comparison with TDS-1 data. The sensitivity
analysis consisted in varying one parameter at a time in
a selected range, keeping all the others fixed.
In Figure 20, we report the left-right (LR) reflectivity
versus the SM as predicted by SAVERS. Here we show that
the SM sensitivity changes according to the surface topography.
Considering a real digital elevation model (DEM) of a
volcanic area in Chad, we have selected two SPs at different
elevations along a TDS-1 track and plotted the reflectivity
as a function of the SM. The corresponding reflectivity over
a flat terrain [the top blue line in Figure 20(a)] is reported
as a reference; the small-scale roughness was set to 1.5 cm
with no vegetation on top. It is possible to observe that the
sensitivity to SM is quite independent of the surface elevation,
though the LR reflectivity gets lower as the elevation
gets higher.
In Figure 21, the sensitivity to forest biomass is compared
for SPs at different heights along a TDS-1 track over the
-158
-160
-162
-164
-166
-168
-170
-172
same mountainous terrain. For ease of reference, the same
geographical locations as in Figure 20 have been selected,
though it is not actually a forested site. The biomass variation
was set in the range [25÷400] t/ha, which is linked to a
variation of the mean diameter at breast height within the
range [7÷21] cm and of the density within [350÷100] trees/
ha through the allometric equations in [64]. The reflectivity
corresponding to a flat terrain is reported as a reference [the
top blue curve in Figure 21(a)]. It can be observed that the
sensitivity to biomass decreases over areas with an important
topography. Since the elevation decreases the reflectivity
value, the coherent scattering may eventually become
weaker than incoherent scattering from the vegetation volume
and underlying soil. In this case, the saturation threshold
is reached, as happens in the monostatic case, and the
sensitivity to biomass may be lost. We remark that SAVERS
is able to model not only the attenuation contributions connected
to the VOD, but also the volume scattering from the
vegetation canopy [38].
y = 17*x - 1.7e+02
Data 1
Linear
43 dB (0 - 0.15)
17 dB (0 - 0.45)
8.6 dB
(0.15 - 0.45)
0 0.05 0.1 0.15 0.2 0.25 0.3 0.35 0.4 0.45
SM
(a)
-163
-164
-165
-166
-167
-168
-169
-170
-171
y = 43*x - 1.7e+02
Data 1
Linear
-160.5
-161
-161.5
-162
-162.5
-163
-163.5
-164
0.05
SM
(b)
0.1
0.15
y = 8.6*x - 1.6e+02
Data 1
Linear
0.15 0.2 0.25 0.3 0.35 0.4 0.45 0.5
SM
(c)
FIGURE 18. The simulated DDM peak values with GARCA as a function of SM. (a) Simulation over the entire SM dynamic range (from [49])
and corresponding mean sensitivity are compared to the sensitivity estimates for (b) small and (c) large values of SM.
30
IEEE GEOSCIENCE AND REMOTE SENSING MAGAZINE JUNE 2022
DDM Peak (dBW)
DDM Peak (dBW)
DDM Peak (dBW)
IEEE Geoscience and Remote Sensing Magazine - June 2022

Table of Contents for the Digital Edition of IEEE Geoscience and Remote Sensing Magazine - June 2022
