With the waters of the Mekong and Tonle Sap lapping the top of their banks

Cambodia looks likely to be hit by widespread flooding in the coming weeks

- particularly with heavy rain in areas of northeast Thailand that drain into

the Mekong. Sarah Stephens discovers how a new application of radar

technology may bring about a change for those worst-affected by the country's

weather.

AGENCIES and NGOs may have contingency plans for

emergencies such as flooding, but getting accurate and reliable information in a

real disaster is always difficult.

Overlaid RadarSat images of the Mekong from Kampong Cham to Neak Luong in Prey Veng show how water has inundated the land at given times. In this particular image, black areas are water that was present during all three radar shots, while red indicates water present during the wet season scans (August and October), but not during the dry season (January).

"If there is a major flood, you often

get a mismatch of information from ground staff," said Mack Ramachandra, of the

Vulnerability Analysis Mapping unit at the World Food Program. "Are sixteen, or

twenty-two, or thirty communes affected? This mismatch happens a lot, and it

happens around the world, not just in Cambodia."

Ground reports have been

the most reliable way for agencies to gauge their target areas for emergency

relief. But now there is new hope for Cambodia in the form of an "eye in the

sky": satellite radar technology.

Since 1997, World Food Program (WFP)

has been working on a pilot program with Canadian-based radar specialists

RadarSat International, and Aruna Technology Ltd, a Cambodian company

specializing in remote sensing and information technology.

The aim of the

program, for which the pilot project is nearing completion after two years'

work, is to create a comprehensive radar image of Cambodia's flood plains at

various times throughout the year, eventually enabling relief agencies to

accurately pinpoint areas most at risk of extreme flooding.

Craig Johnson

of RadarSat International is convinced that radar satellite technology is

indispensable for disaster management in Cambodia.

"One of RadarSat's

most important features is that it can capture images through clouds, which is

very useful in a country like Cambodia," he explained. "Optical satellites need

to use daylight to obtain images. Radar's microwaves penetrate even thick cloud

cover and can be used in darkness. The satellite also has a steerable sensor,

which means we can take images at very short notice - we don't have to wait for

the satellite to be in the right place above the earth."

Because of the

way that the radar beam is reflected from objects on the earth, any area covered

by water shows up a solid black color on the final image. By color coding images

taken at different times of the year, analysts at Aruna Technology can build up

a highly accurate map of the extent of flooding at different times during the

rainy season.

When this is combined with information from gauge stations

along the Mekong River, which measure the height of the river at specific times

and locations, it becomes possible to predict how far the floods will spread

when the river rises.

With this information, say Aruna and Radar Sat, it

will be possible to predict which areas will be affected before the flooding

even occurs, just from noting the height of the Mekong, and comparing those

figures to the radar maps of previous flooding.

The pilot program

covered two specific areas along the Mekong, the first from Kratie to Kampong

Cham, the second from Kampong Cham to Prey Veng. Images were taken in August,

October and December 1997, and in January 1998.

When the images are

overlaid (as shown in the accompanying illustration) they provide a detailed

picture that shows clearly to what extent water has inundated the land at given

times. In this particular image, black areas are water that was present during

all three radar shots, while red indicates water present during the wet season

scans (August and October), but not during the dry season (January). These areas

are therefore subject to seasonal inundation.

WFP are still waiting for a

final analysis and interpretation of the images from Aruna Technology, but so

far, said Ramachandra, the results are encouraging.

"We already have a

database [at WFP] which shows various socio-economic indicators across the

country, which we can display on our own maps," he said, "but combined with the

satellite images and the ground information we can gain a much more accurate

picture of what is happening in the affected areas."

Radar Sat has

already been used successfully in disasters across the world, including

Hurricane Mitch, which devastated Central America late 1998, and the Yangtze

River flood in China in July and August 1998.

The technology was

originally conceived to chart iceflows in the Canadian Arctic, and, as Andrew

McNaughton, Director of Aruna Technology pointed out, the applications of radar

technology are far more wide-ranging than just flood mapping.

"We are

developing the capability to monitor the progress of rice crops," he said, and

will soon be able to predict yields for any given area down to commune level.

The technology relies on radar's ability to sense the texture and extent of

plant cover in the wet rice field, which can be interpreted to give an

indication of the condition of the crop."

Radar images taken by the team

also clearly show fish traps and other activity in the Tonle Sap, as well as

road networks and forest coverage, urban development and other land use

information. Objects as small as eight meters across can usually be

distinguished.

One downside to remote sensing technology, including

radar, is its initial cost. While McNaughton and Johnson are keen to point out

that images bought in bulk are cheaper, a single image can still cost between

$4,000 and $6,000, plus processing costs, which may be way over the budget of

some of the smaller agencies who could benefit from the technology. In addition,

if an image is needed immediately (Radar Sat can deliver images in just over one

day) the price rises accordingly.

"NGOs who want to use the same image

can group together to share costs," said McNaughton.

But for Ramachandra

at WFP, there's a delicate balance between the cost of the radar imagery and the

amount of money needed for an emergency operation.

"If it is going to

cost $5,000 for a relief operation, do I want to spend $20,000 on images from

radar?" he mused.

Of course, one thing for agencies to bear in mind is

that once an image is bought, its data remain pertinent for many years, so the

long-term payoff on the initial investment can be large. As part of its business

strategy of helping Cambodian institutions develop their own capacity, Aruna is

developing plans to assist them to acquire nationwide coverage for general

use.

In theory, a permanent reference library could be created by mapping

all areas of the Mekong during each season - images that could be used in

development, forestry, agriculture, and fisheries, as well as emergency

work.

But this would all be several years down the line. For now, Aruna

and RadarSat are concentrating on the final results of the pilot

program.

"To complete the pilot project we need to do an assessment, meet

with the Cambodian agencies who would have use for the data, fine tune the cost,

define the players, then find the funding," said McNaughton.

He is

confident, though, that there will be a real use for the project's results, as

sustainable development in Cambodia continues to require ever-better information

for planning and decision making.