The Full Picture

[

] 238

Rapid biological and ecological

assessments using Earth observation data

Douglas M. Muchoney, Group on Earth Observations

arth observation data including satellite, aerial and ground

data have an extensive history of use to characterize and

monitor marine, coastal, freshwater and terrestrial environ-

ments. Rapid assessments are undertaken for a whole host of

reasons, including protected areas system design and gap analy-

ses, protected areas assessments, threat assessments, management

planning, and biological and ecological inventories.

Such assessments are conducted across the entire array of the Earth’s

ecosystems including marine parks and corals reefs, coastal zones,

wetland and forests. Rapid assessments are undertaken when there is a

need to quickly obtain information on biological, ecological and envi-

ronmental status over large areas, and where there are constraints on

resources or access issues such as harsh or dangerous terrain.

Rapid assessments take advantage of the synoptic view that satellites

provide, in addition to other geospatial data such as soils, geology and

terrain data (often satellite-derived as well). They often use aerial data

such as photography and videography, and collect field data to under-

stand the relationship between what is observed from space

and the air, to actual ground conditions. The field data are

also used to ‘teach’ or ‘train’ computer pattern recognition

algorithms to recognize different types of vegetation. The

pattern recognition software can be quite straightforward,

involving just a satellite image interpreter that uses colour,

tone, pattern and texture to distinguish between different

features of the landscape, while others use sophisticated

decision tree and neural network algorithms that are

designed to function as the human mind does.

In this sense, rapid assessments incorporate many of

the aspects of GEO: they require satellite data for synop-

tic views of the landscape, they use intermediate aerial

and high-resolution satellite image data, they integrate

remote sensing and other geospatial data, and they

employ field data collection to understand the satellite

and aerial data. The data that are used include soils,

geology, digital terrain, climate and hydrology. There is a

Landsat Thematic Mapper Data 2002 (left) and 2004 (right)

The data enabled the researchers to map the extent and condition of the

groundwater swamp forests of the Prey Long region, as well as its alteration over time

Source: Landsat satellite data, courtesy US Geological Survey

OCIETAL

ENEFIT

REAS

– E

COSYSTEMS