drainage pattern are closely related to the lithology, i.g.
resistence to erosion, permeability of the ground, slope
angle, and tectonic control.
One reason why imaging radar is useful for mass move-
ments applications is ist sensitivity to surface roughness
and slope. The intensity of radar backscatter strongly
depends on the local slope and is also affected by smal-
ler-scale surface roughness (Pettengil et al., 1986). The
roughness of a surface serves as an important attribute in
The change detection method minimises the impact of
target variables such as soil texture, roughness, and
vegetation, because these tend to change slowly, if it all.
Also fortunatly for many hydrologic applications, the
changes in soil moisture may be more important than the
actual monitoring (Engman 1991; Lichtenegger 1992).
Spaceborne SARs are well-suited for soil moisture inve-
stigations over large areas. For instance, a SAR flown
800 km altitude with incidence angles of 17° to 23° would
terrain analysis and is often closely related to the cover a swath of approximatly 100 km. 5
(underlying) geological substrate. Roughness may also
be caused by weathering processes, soil composition, or =
vegetation associations. In areas where the surface is 4. CONCLUDING REMARKS
unvegetated and dry, in arid or semiarid regions, deep
penetration is possible. On the other hand the penetration In Table 2 a summary of the usefulness of satellite re- G
capabilities in densily vegetation-covered regions are mote sensing in the different phases of disaster mana- PI
hampered by surface and volume scattering. It depends gement for flooding, earthquakes, volcanic eruptions, and "R
on what one is looking for, either to perform topographic landslides is given. From this table it can be concluded as
analysis or to investigate vegetational effects the obtained that most promising results can be expected in the fields ES
results will be completely different. The depth of penetra- of. volcanic eruptions and flooding, as both types of [C
tion of electromagnetic waves normally incident from air disasters result in features that are clearly recognisable LM
or space onto the earth increases with longer wave- with the use of satellite imagery. Earthquakes and lands- Sc
lengths and also with a decrease in the attenuation losses lides generally result in damages to objects that are too IG
and volume scattering within the shallow subsurface The small to recognise on the current imagery. FI
attenuation in the soil is governed by soil moisture (one Table 3 lists the current satellite remote sensing data that Si
per cent effectively rules out any deeper penetration) and could be used in disaster management. "Sr
soil texture. (Drury, 1993; Carver et al., 1987). Lm
SAR is especially sensitive to the presence of water, 1
3
gt
Table 2 Usefulness of Satellite Remote Sensing for Disaster Management: 2
s
Disaster type Disaster prevention Disaster preparedness Disaster relief
Volcanism T Tt ++ Te
Earthquakes + - 0 va
Landsliding 0 + +
Flooding ++ ++ ++ | Ge
ge
++ = very useful, + = useful, = of limited use, - = not useful
Di:
Table 3 The current satellite remote sensing data that could be used in disaster management. Cr
Disaster type Disaster prediction Disaster preparedness Disaster relief M
Volcanism TM/ SPOT (radar)/ TM/ NOAA TM/ SPOT/ GOES/ TOMS/ La
ERS/ JERS ERS/ JERS
Earthquakes TM/ SPOT/ ERS/ JERS - TM/ SPOT/ ERS/ JERS Rc
Landsliding SPOT/ ERS NOAA/ ERS/ JERS TM/ SPOT/ ERS/ JERS
Flooding TM/ SPOT/ ERS/ JERS NOAA/ Meteosat TM/ SPOT/ ERS/ JERS | Fic
Aq
either in the form of soil moisture, liquid water in a sno- D
wpack, or vegetation moisture. The extraction of hydro Finally, the following conclusions can be stated: So
information from SAR imagery of terrain surfaces is a * The existing tools can generally be considered ade- [
difficult task, because the image intensity is a complex quate. Currently SPOT and Landsat TM are the most Sn
function of many radar and target parameters. These may used systems. |.
include wavelength, incidence angle polarisation, soil * Temporal resolutions (and spatial resolution) should Sn
moisture, vegetation cover, surface roughness, slope be improved. There is however, a clear need for cer-
aspect, and temporal cover, e.g. a snow coverage. Multi- tain types of disasters (earthquake, landslides) to ha-
polarised data can, in principle, be used to separate sur- ve stereoscopic data with a larger spatial resolution. Un
face roughness effects from soil moisture effects. The * In many applications weather condition are the most
dielectric properties of soil are essential in determining important drawback. In the near future, however, it is Ke
microwave backscattering and absorbtion by soil. expected that many applications will derive from the d
An addional approach for using soil moisture data derived use of ERS and JERS data, especially in areas where HA
with microwave approaches is through change detection. EV
DR
PL
42
International Archives of Photogrammetry and Remote Sensing. Vol. XXXI, Part B6. Vienna 1996
