Use of impervious surface data obtained from remote sensing in distributed hydrological modeling of urban areas - Urban Remote Sensing: Monitoring, Synthesis and Modeling in the Urban Environment - page 257

accurate than high-resolution remote sensing for modeling the
spatial distribution of runoff at a 30m cell size. However, if
high-resolution data are not available or if resources are limited,
the technique may be very useful to account for spatial variation
in runoff.

better discrimination between spectrally similar impervious and
non-impervious cover types, both at high resolution (Franke
et al ., 2009) and medium resolution (Weng, Hu and Lu, 2008;
Demarchi et al ., 2010), leading to further improvement in the
accuracy of impervious surface mapping. The increasing spatial
resolution of hyperspectral sensors will be important as well for
mapping of spatially heterogeneous urban land-cover. For the
modeling of urban rainfall-runoff many challenges lie ahead.
Methodologically, a strong focus should be given to the coupling
of hydrological models for the estimation of runoff with hydraulic
models for river and especially sewer flow. Also interaction of
hydrological processes at the land surface, sewer and drinking
distribution networks with the groundwater needs conceptual
model improvements. Effects of climate change on hydrological
processes in urban environments as well as demographic and
economic changes will impact land and water use. New study
and predictions approaches for adapting land and water manage-
ment will have to be developed. Finally, the challenge remains, as
already identified in the NEMO project in the 1990s, to translate
scientific research results into actual urban water and land man-
agement or policy. Collaboration over the borders of scientific
disciplines and with land and water policy/management will thus
be required.

Conclusions

Until recently the relationship between flood conditions and the
spatial distribution of urban development has been poorly stud-
ied, partly because of the traditional use of lumped models in
urban hydrological modeling. Presently, distributed hydrological
models are becoming more popular because of their advantage
of taking the spatial dynamics of hydrological processes explicitly
into account. The fact that the spatial distribution of physi-
cal characteristics in hydrological models is mostly described
by raster data makes the link to remotely sensed data obvious
and highly interesting. One of the most important factors influ-
encing runoff in urbanized areas is the extent and the spatial
distribution of impervious surfaces within the catchment. In
recent years, much research in urban remote sensing has been
devoted to the development of methods for accurate mapping of
impervious surfaces. Several studies addressed detailed impervi-
ous surface mapping from high-resolution remote sensing data.
Other research focused on fractional mapping of impervious sur-
face cover from medium-resolution data at the sub-pixel scale,
offering the possibility to map impervious surface cover for large
areas at reasonable cost.
In this study the impact of different methods for characterizing
the distribution of impervious surfaces on the estimation of peak
discharges was assessed for a strongly urbanized watershed in the
Brussels Capital Region. The study shows that estimates of aver-
age imperviousness derived from satellite data may strongly differ
from expert knowledge on the imperviousness of urban land-use
classes documented in the literature and may lead to substantially
different estimates of discharge at catchment level. The study also
demonstrates that use of spatially detailed information on imper-
viousness obtained from high-resolution remote sensing leads
to substantially higher estimates of peak discharge, compared to
the use of average levels of imperviousness for different types
of land-use, which can be considered the standard approach in
hydrological modeling. The spatial variation in runoff and the
connectivity between cells with high runoff coefficients in partic-
ular, proves to be a major factor influencing estimated discharge
volume. Sub-pixel estimation of land-cover class proportions
based on Landsat ETM

Acknowledgments

The research presented in this chapter is partly funded
by the Belgian Science Policy Office in the frame of the
STEREO programme - project SR/00/02 (SPIDER). J. Chor-
manski acknowledges the support of the Research in Brussels
Programme for a postdoctoral fellowship.

References

Abbott, M., Bathurst, J., Cunge, J. et al . (1986) An introduction
to the European Hydrological System - Systeme Hydrologique
Europeen, ''SHE'', 2: Structure of a physically based, dis-
tributed modeling system. Journal of Hydrology , 87 , 61 - 77.
Anderson, D.G. (1968) Effects of urban development on floods
in northern Virginia. US Geological Survey Water-Supply Paper
2001-C.
Anderson, K. and Croft, H. (2009) Remote sensing of soil surface
properties. Progress in Physical Geography , 457 - 473.
Arnold, C.L., Jr. and Gibbons, C.J. (1996) Impervious surface
coverage: the emergence of a key urban environmental indi-
cator. Journal of the American Planning Association , 62 (2),
243 - 258.
Arnold, C.L., Jr., Crawford, H.M., Gibbons, C.J. and Jeffrey, R.F.
(1994) The use of Geographic Information System images as a
tool to educate local officials about the land use/water quality
connection, in Proceedings of the Watershed '93 Conference ,
Alexandria, Virginia, March 1993, pp. 373 - 377.
Batelaan, O. and De Smedt, F. (2007) GIS-based recharge esti-
mation by coupling surface-subsurface water balances. Journal
of Hydrology , 337 (3 - 4), 337 - 355.

data having the same resolution as the
model's grid cell size (30 m) produces less structural detail, yet
spatial patterns of imperviousness are well characterized. The
higher level of generalization results in a smoothing of the runoff
coefficient pattern and leads to slightly lower peak discharge esti-
mates compared to the use of high-resolution impervious surface
maps as input for the modeling. However, if high-resolution
data are not available or if resources are limited, the technique
may be an interesting alternative for high-resolution impervious
surface mapping in rainfall-runoff modeling within urbanized
catchments.
A lot of open research issues are recognized, which should
allow future improvements in both impervious surface mapping
as well as runoff modeling. New hyperspectral sensors allow a

+