nt, for a
or other
solution
several
quently
ling. In
ally the
ich the
e flood
re less
ause of
images
reduces
egional
logical
re their
ground
ry, and
quired
RSAT
sensor
Earth-
atively
ig the
if the
useful
A big
a few
ly by
sible.
nental
A-14,
rth on
nes a
5, the
gical
twice
Ittime
nages
ver is
oring
partly
od of
dent)
SAT.
m the
nique
t, but
this can cause additional difficulties in the further analysis. SAR
sensors can penetrate clouds, but with a return period of more
than five weeks, two of them are only used for flood monitoring
by chance. RADARSAT is able to receive images over the same
area within a few days, but here the problem of fast availability
and costs of images aeed further considerations.
Availability of images. In flood monitoring, near real-time data
are needed to enable disaster mitigation processes.
Meteorological satellites have virtually continuous data
acquisition facilities at numerous locations in the world, and
data can be reached fast through computer networks or via
direct satellite receiving. In case of downloading from computer
networks, only limited preprocessing is carried out operationally
at the receiving stations, therefore the user has to accomplish
atmospheric correction, geometric rectification and other
processes needed for the analysis. In the case of high-resolution
images (optical and radar), the satellites are not acquiring data
on a continuous basis. For most of the radar images orders have
to be submitted to the operating organizations beforehand,
which in most flood events is not possible because of the fast
occurrence of flooding. Orders usually have to be submitted a
few days to a few weeks before the planned acquisition time.
Another issue is the availability of data acquired previously and
stored in digitized form. The archiving of data once collected
from satellite platforms is a problematic issue for organizations
operating the satellite systems. Storage facilities and the
archiving process need a tremendous amount of storage capacity,
labor and investment. Different organizations at different
locations archive the data according to different guidelines,
which makes it troublesome to inquire about and check the
availability of images taken at certain locations, certain time and
with a certain sensor. In addition, only a part of the received
information is stored for further utilization, a significant part is
not archived after reception.
Costs of data. Costs of data are always an important, and often
a decisive issue in all satellite applications. The data from
meteorological satellites are virtually free, in most cases costs
only apply for copying and delivering the images to the user
(Table 1). This is in contrast with the high-resolution optical
and radar images, which usually need a larger budget on the
user side, especially when using multiple sets of images for a
detailed analysis. Fresh images tend to be the most expensive,
and data form the archive can be often obtained with discount.
In case of flood monitoring, near real-time analysis faces the
problem of funding, while scientific research applications are
able to utilize archived data for a much lower cost.
As an outlook, Table 2 shows different applications of remote
sensing in the field of hydrology. A wide range of sensors
operating onboard satellites at present are used, and others are
planned for launch in the near future. However, it is necessary
to mention that none of the sensors presently in operation were
designed directly to measure hydrological parameters above
land surface, thus the spatial resolution, return period, and other
characteristics of the sensors are not ideally adjusted to
applications such as flood monitoring.
ASSASSMENT OF INUNDATED AREAS
In monitoring floods, a fundamental problem is the collection of
the available satellite images. In most cases the images arrive at
the place of processing in a few days to a few weeks after the
acquisition, which may be sufficient for scientific research but
may not be sufficient for urgent emergency and disaster
Variable Currently Available Planned Satellites/Sensor|
Satellites/Sensor Systems Systems
AVHRR
LANDSAT MODS
SPOT A
ATSR
Snow cover | SSM/I MIMR
ATSR/ERS-1 SSM/A
ATSR/ERS-2 ICESTAR
RADARSAT
ALT
Water SSALT
elevatión TOPEX-POSEIDON RA
GEOSAT
TRMM
Rain GOES aMSU
rate/amount| SSM/I VISSR
GOMS
AMI/ERS-1
Soil ANU/ERS-2
moisture JERS-1 HYDROSTAR
SSM/I
RADARSAT
CERES
Solar POLDER
adiation METEOR/SLARAB SCARAB
GERBI
MODIS
AVHRR AVERR
Surface ATSR/ERS-2 ASTR
albedo GOES POLDER
GOES
VISSR
MODIS
AVHRR
AVHRR ATSR
Land LANDSAT AATSR
cover/use/ts | SHOT OCTS/GLI
index ATSR/ERS-2 POLDER
ASTER
TM
LANDSAT
SPOT
Flood AVHRR AVHRR
monitoring | ERS-1/ERS-2 TM
JERS-1
RADARSAT
MODIS
AVHRR
Surface AVHRR OCTS/GLI
temperature| LANDSAT TM GOES
MVIRI/SEVIRI
GOMS-BTVK
Table 2: Some applications of sensors for measuring hydrological
variables (after Engman, 1996)
mitigation planning. Meteorological satellites can supply near
real-time (or even real-time) images over the critical sections of
rivers, but the applications of these images are often restricted
by significant cloud cover. Thus, the near real-time flood
monitoring faces a basic technological problem of receiving
data in time. In most flood events, only a few images are
available for the researchers: a high-resolution optical image,
used as a reference and taken usually months before the event,
meteorological images with different cloud coverage and
possibly radar images.
Availability of images. When only optical sensors are available
for the analysis one has to be fortunate in acquiring enough
number of images without significant cioud cover. During the 6
May 1990 flooding in the delta of the Krishna river in south
India IRS-1A, NOAA and Landsat data were used for
Intemational Archives of Photogrammetry and Remote Sensing. Vol. XXXII, Part 7, Budapest, 1998 781
a
