Forests for People

[

] 296

developed the original software to detect the suspected

logging areas routinely and smoothly. Functions of

the software include making multi-temporal colour

composite images, cutting out certain areas from the

colour composite image, and extracting those areas as

GeoTIFF or KML formats.

Estimating forest biomass in Indonesia

Since the 1990s, Indonesia has used optical satellite data

(Landsat) for forest monitoring, but cloud cover has been

information. If large areas are deforested during the period between

the two sets of observation data, the colour composite image would

show an area of deforestation as darker than a normal forest area.

Furthermore, new deforestation areas are shown as a dark reddish

colour, while old deforestation areas are shown as dark brown.

In other words, illegal logging is probably under way in the areas

shown as dark reddish in the colour composite image.

The colour composite image was divided into several scenes

and delivered to interpreters of IBAMA and DPF in order to

detect the dark reddish suspected logging areas. There are several

interpreters to check the colour composite images covering the

whole area of Brazilian Amazon forest. The divided scenes can be

extracted as GeoTIFF or KML format so that the interpreters can

easily overlay the images on the other optical satellite images on

their GIS or even using Google Earth. Needless to say, GIS helps

the interpreters identify the position of suspected logging areas.

When the interpreters find such areas, the DPF officials rush to the

area to prevent further illegal logging and protect the forest. The

rapid delivery, process and analysis of PALSAR data is the key to

preventing illegal logging. This has only been made possible with

ScanSAR data because it can cover the whole of Brazil within five

or six days. While ground resolution of ScanSAR data is only 100

m, speed of delivery and the extent of coverage are more important

for this project.

To complete the project, RESTEC transferred the required analysis

techniques through training. As radar satellite data is very different

from optical data, RESTEC started the first training with very basic

radar theory and handling techniques. Then, more advanced tech-

niques such as change detection, multi-temporal composite, visual

interpretation and so on were transferred. In parallel, RESTEC

Colour composite image

PALSAR ScanSAR strip data

Colour composite images can be made from multi-temporal strip data. New

deforestation areas are visible as a reddish dark colour and old deforestation

areas are dark brown

The original PALSAR ScanSAR strip data is a black and white

image, with a single strip covering an area 350 km wide and

3,000 km long