performance than scenario 1 with an 11 to 27 percent im-
provement.
Based on these results, it can be inferred that texture fea-
tures can increase the accuracy of conductivity estimation for
all
AOIs
. According to Equation 6 roughness information and
RMS
slope of surface can affect the backscattering data. We
believe that the texture features used in scenario 3 provide
useful information related to surface roughness which im-
proves the estimate of dielectric constant and thus observed
conductivity.
We should note that because our method is applied to such
a different environment (constructed earthen levees, having
steep slopes) from other studies of soil moisture estimation
from radar, and furthermore uses a shorter wavelength (most
previous work used L or C band), we do not attempt here to
compare our results to them.
Finally it is worth mentioning that our algorithm is trained
based on the training samples from the areas shown in Figure
2. If a region on a levee covers different grass or canopy, we
need to have enough training samples from the new area and
train the machine learning algorithm for that new area.
Conclusion
In this work, a machine learning approach based on
BPNN
and
WBNN
is investigated for estimation soil electrical con-
ductivity on an earthen levee system using X-band
SAR
. The
algorithm was examined in four areas of study with different
vegetation covers, using three scenarios based on the extract-
ed feature sets used. TanDEM-X dual-polarization
SAR
data
is used along with
DEM
data obtained from lidar to compute
local incidence angles. We used reference data soil electrical
conductivity measurements for training and validation of the
algorithm. The results show that scenario 3, which includes
all backscatter and texture features, has better correlation
performance than the others with around 11 to 27 percent im-
provement compared to scenario 1 (backscatter features only)
and 1 to 17 percent improvement over scenario 2 (backscat-
ter plus window statistics). This result shows that estimat-
ing conductivity based on only backscatter features cannot
provide suitable estimation for this environment and we need
to include texture information. The results also show that the
WBNN
performance is very similar to that of the
BPNN
. Based
on these results, our approach shows some promise for use in
a levee health monitoring system based on X-band
SAR
.
References
Akbar, M.A., A.L. Kenimer, S.W. Searcy, and H.A. Torbert, 2005. Soil
water estimation using electromagnetic induction,
American
Society of Agricultural Engineers
, 48:129–135.
Attema, E.P.W., and F.T. Ulaby, 1978. Vegetation modeled as a water
cloud,
Radio Science
13:357–364.
Baghdadi, N., M. Aubert, and M. Zribi, 2012. Use of TerraSAR-X data
to retrieve soil moisture over bare soil agricultural field,
IEEE
Geoscience and Remote Sensing Letters
, 9(3):512– 516.
Bindlish, R., and A.P. Barros, 2001. Parameterization of vegetation
backscatter in radar-based, soil moisture estimation,
Remote
Sensing of Environment
, 76(1):130–137.
Burrus, C.S.,R.A. Gopinath, and H. Guo, 1997.
Introduction to
Wavelets and Wavelet Transforms: A Primer
, Prentice-Hall,
Upper Saddle River, New Jersey.
Cressie, N.A.C., 1990. The Origins of Kriging,
Mathematical Geology
,
22:239–252.
Cui, M., S. Prasad, M. Mahrooghy, J.V. Aanstoos, M. Lee, and L.M.
Bruce, 2007. Decision fusion of textural features derived from
polarimetric data for levee assessment,
IEEE Journal of Selected
Topics in Applied Earth Observations and Remote Sensing
,
133(5):575–587.
De Roo, R.D., Y. Du, F.T. Ulaby, and M.C. Dobson, 2001. A semi-
empirical backscattering model at L-Band and C-Band
for a Soybean canopy with soil moisture inversion,
IEEE
Transactions on Geoscience and Remote Sensing
, 39(4):864–872.
Dobson, M.C., F.T. Ulaby, M.T. Hallikainen, and M.A. El-Rayes, 1985.
Microwave dielectric behavior of wet soil, Part II: Dielectric
mixing models, I
EEE Transactions on Geoscience and Remote
Sensing
, GE-23:35–46.
Du, Y., F.T. Ulaby, and M.C. Dobson, 2000. Sensitivity to soil moisture
by active and passive microwave sensors,
IEEE Transactions on
Geoscience and Remote Sensing
, 38(1).
Dubois, P.C., J.J. van Zyl, and T. Engman, 1995. Measuring soil
moisture with imaging radars,
IEEE Transactions on Geoscience
and Remote Sensing
, 33(4):896–904.
Düring, R., F. Koudogbo, and M. Weber, 2008. TerraSAR-X and
TanDEM-X: Revolution in Spaceborne Radar, ISPRS
Archives
for Photogrammetry, Remote Sensing, and Spatial Information
Sciences
, XXXVII.
Fausett, L.V., 1993.
Fundamentals of Neural Networks: Architectures,
Algorithms and Applications
, Prentice Hall.
Frison, P.L., E. Mougin, and P. Hiernaux,1997. Observation and
interpretation of seasonal ERS-1 wind scatterometer data
over Northern Sahel (Mali),
Remote Sensing of Environment
,
63:233–242.
Hajnsek, I., T. Jagdhuber, H. Schon, and K.P. Papathanassiou,2009.
Potential of estimating soil moisture under vegetation cover by
means of PolSAR,
IEEE Transactions on Geoscience and Remote
Sensing
, 47(2):442–454.
Hallikainen, M.T., F.T. Ulaby, M.C. Dobson, M.A. El-Rayes, and
L.K. Wu, 1985. Microwave dielectric behavior of wet soil-Part
1: Empirical models and experimental observations,
IEEE
Transactions on Geoscience and Remote Sensing
, 23(1):25–34.
Huth, N.I., and P.L. Poulton, 2007. An electromagnetic induction
method for monitoring variation in soil moisture in agroforestry
systems,
Soil Research
, 45:63–72.
Jin, N., and D. Liu, 2008. Wavelet basis function neural networks
for sequential learning,
IEEE Transactions on Neural Networks
,
19(3):523–528.
Kim, S., T. Leung, J.T. Johnson, H. Shaowu, J.J. van Zyl, and
E.G. Njoku, 2012. Soil moisture retrieval using time-series
radar observations over bare surfaces,
IEEE Transactions on
Geoscience and Remote Sensing
, 50(5):1853–1863.
Kim, Y., J.J. van Zyl, 2009. A time-series approach to estimate soil
moisture using polarimetric radar data,
IEEE Transactions on
Geoscience and Remote Sensing
, 47(8):2519–2527.
Kseneman, M., D. Gleich, and Z. Cucej, 2011. Soil moisture
estimation using high resolution Spotlight TerraSAR-X data,
IEEE Geoscience and Remote Sensing Letters
, 8(4):686–690.
Mahrooghy, M., J. Aanstoos, K. Hasan, S. Prasad, and N.H. Younan,
2011. Effect of vegetation height and volume scattering on
soil moisture classification using synthetic aperture radar
(SAR) images,
Proceedings of the 2011 IEEE Applied Imagery
Pattern Recognition Workshop (AIPR)
, pp. 1–5. doi:10.1109/
AIPR.2011.6176375,: 1–5.
Mattia, G., V.R. Satalino, N. Pauwels, and A. Loew, 2009. Soil
moisture retrieval through a merging of multi-temporal L-band
SAR data and hydrologic modelling,
Hydrology and Earth
System Sciences
, 13:343–356.
Miller, P.E., J.P. Mills, S.L. Barr, S.J. Birkinshaw, A.J. Hardy, G. Parkin,
and S.J. Hall. 2012. A remote sensing approach for landslide
hazard assessment on engineering slopes,
IEEE Transactions on
Geoscience and Remote Sensing
, 50(4):1048–1056.
Morris, E.R., 2009. Height-above-ground effects on penetration depth
and response of electromagnetic induction soil conductivity
meters,
Computers and Electronics in Agriculture
, 68(2):150–156.
Morvan, A., M. Zribi, N. Baghdadi, and A. Chanzy. 2008. Soil
moisture profile effect on radar signal measurement,
Sensors
,
8(1):256–270.
National Climatic Data Center, 2009.
National Overview
.
518
July 2016
PHOTOGRAMMETRIC ENGINEERING & REMOTE SENSING
