004
ERS
'eral
iken
rom
| the
for
ined
; by
n on
f the
This
how
oned
t the
nties
r set
ps at
d by
ange
rved
n by
most
point
er at
may
water
eing
ering
f the
may
o the
es of
«tent.
International Archives of the Photogrammetry, Remote Sensing and Spatial Information Sciences, Vol XXXV, Part B7. Istanbul 2004
By incorporating the ERS objective, the subsequent parameter
constrain becomes visible in some areas.
S. CONCLUSION
The findings in this flood propagation study showed the
equifinality of roughness coefficients and outlined the need for
multi-response evaluation. Most importantly, it was shown that
simulations calibrated with radar data performed almost equally
well than the models only conditioned on ground data. The
main difference between both calibration methodologies can be
related to the increased fuzziness of earth observation data that
leads to larger prediction uncertainties. Due to this redundancy,
the responses of models that were initially conditioned using
measured high water marks could not be significantly
constrained with synchronically obtained radar observations.
This does not mean that on different sites with sparse ground
data sets, the constrain could not become significant. On our
test site, however, a significant constrain was only achieved
with radar data sets obtained several hours before peak flow
occurred. This suggests that in order to become complementary
to existing ground data, the radar coverage should be different
in time and/or space from the point data sets. This approach
could also help addressing the well-known problem of changing
roughness values with increasing water levels. However, if the
time interval between available data sets is too long, this may
lead to the rejection of all model simulations. Therefore, it will
be interesting to investigate whether additional data sets of
different flood events will further constrain the plausible
parameter sets or, in contrast, will lead to the rejection of all
simulations.
Figure 7. Comparison of the “best” simulation (based on radar
observation) and the corresponding Envisat derived
flood area
It has also been pointed out in this study that the application of
à fuzzy rule based calibration technique, along with a
generalized likelihood ^ uncertainty estimation (GLUE)
procedure, constitutes a valuable approach in inundation
modelling. Fuzzy performance measures are perfectly suited for
radar data with no knowledge of the error structure. Dealing this
way with the most important sources of uncertainties could
ultimately lead to an increase of confidence that flood managers
will have in the simulation results.
357
6. REFERENCES
Aronica, G., Hankin, B., Beven, K.J., 1998. Uncertainty and
equifinality in calibrating distributed roughness coefficients in a
flood propagation model with limited data. Advances in Water
Resources, 22(4), pp. 349-365.
Aronica, G., Bates, P.D., Horritt, M.S., 2002. Assessing the
uncertainty in distributed model predictions using observed
binary pattern information within GLUE. Hydrological
Processes, 16, pp. 2001-2016.
Beven, KJ, 1993. Prophecy, reality and uncertainty in
distributed hydrological modeling. Advances in Water
Resources, 16(1), pp. 41-51.
Biggin, D.S., Blyth, K., 1996. A comparison of ERS-1 satellite
radar and aerial photography for river flood mapping. Journal
of the Chartered Institute of Water Engineers and Managers,
10, pp. 59-64.
De Roo, A., Van Der Knijff, J., Horritt, M., Schmuck, G., De
Jong, S., 1999. Assessing flood damages of the 1997 Oder flood
and the 1995 Meuse flood. Proceedings of the 2"^ International
Symposium on Operationalization of Remote
Enschede, The Netherlands.
Sensing,
Freer, J.E., McMillan, H., McDonnell, J.J., Beven, K.J., 2004.
Constraining Dynamic Topmodel responses for imprecise water
table information using fuzzy rule based performance measures.
Journal of Hydrology,291(3-4), pp. 254-277.
Gupta, H.V., Sorooshian, S., Yapo, P.O., 1998. Toward
improved calibration of hydrologic models: Multiple and
noncommensurable measure of information. Water Resources
Research. 34, pp. 751-763.
Henry, J.-B., Chastanet, P., Fellah, ‘K., Desnos, Y.-L., 2003.
ENVISAT Multi-Polarised ASAR data for flood mapping.
Proceedings of IGARSS'03, Toulouse, France.
Horritt, M.S., Mason, D.C., Luckman, A.J, 2001. Flood
boundary delineation from Synthetic Aperture Radar imagery
using a statistical active contour model. International Journal of
Remote Sensing, 22(13), pp. 2489-2507.
Horritt, M.S., Bates, P.D., 2002. Evaluation of 1D and 2D
numerical models for predicting river flood inundation. Journal
of Hydrology, 268, pp. 87-99.
Pappenberger, F., Beven, K.J., Horritt, M., Blazkova, S., 2004.
Uncertainty in the calibration of effective roughness parameters
in HEC-RAS using inundation data. Journal of Hydrology, in
press.
Romanowicz, R., Beven, K.J., 2003. Estimation of flood
inundation probabilities as conditioned on event inundation
maps. Water Resources Research, 39(3): p. art. no.-1073.
7. ACKNOWLEDGEMENTS
This study is supported by the ‘Ministère Luxembourgeois de la
Culture, de l'Enseignement Supérieur et de la Recherche’. The
authors would like to thank Mr. Jean-Paul Abadie at the French
Space Agency (CNES) for supporting this research project.
