The
analysis of two Landsat Satellite images separated by 15
years is an example of how satellite imagery can contribute
to an environmental audit.
The imagery shown here is Landsat ETM (15m pixel size, Date:
1999) and Landsat MSS (60m pixel size, Date: 1984).
Using
an image difference technique, a coastal change map can be
generated highlighting the transformation of the
coastal area with regard to new vegetation growth and
sea level changes.
Using a classification process on the recent imagery, a
Land Cover Map was created which gives an indication of the
land use categories in the region. Areas likely to be vulnerable
to the environmental impact of commercial activity can be
identified. This is necessary to grade sensitivity and help
establish an emergency contingency plan to deal with potential
degradation and prevent irreversible coastal damage.
Satellite imagery such as this provides a useful means to
monitor river courses, coastlines and mud-banks which can move
hundreds of metres each year. The above images of Bangladesh
illustrate the use of Thematic Mapper (TM) data to study seasonal
flood patterns. The image on the left was captured in the January
dry season, and shows the normal river course at this time
of year. September is the wet season, and the right-hand image
shows the extents of the same river in October, after rainclouds
have cleared. Inundated areas are shown in blue and black,
with this band combination of RGB 453.
The image above left
is again from the Landsat Thematic Mapper (TM) sensor, over
part of the Severn river in the western UK.
The Synthetic Aperture Radar (SAR) image on the right shows
the utility of shallow-angle Radarsat data in delineating
the extents of river flood, as defined by very dark areas.
This
type of information can be used to grade the risk of flooding
to areas containing buildings, as required by the insurance
industry, for example. Radar data also has the great advantage
of being weather independent, thus allowing data collection
during flood peaks, when cloud-cover is inherently widespread.
A Temporal Difference Image (TDI) is
produced by processing two images recorded on different dates,
one before and one after a 'change event'. In addition to earthquakes,
the process can be used, for example, to assess war damage,
movement of military resources, or damage from other forms
of natural disaster.
TheTemporal Difference Image on the right was created
from two IRS optical satellite images, the first taken on
the 7th August 1999, and
the second taken nearly seven weeks later on the 24th of
September. Two extracts are displayed below, showing the destruction
in
towns to the west of Izmit, following a devastating earthquake
on the 17th of August. Extensive areas of building debris
can be identified, along with severe urban damage in central
locations.
There is also some coastal flooding due to land subsidence
apparent on the lower extract.
