GIS/RS
The Geographic Information Systems and Remote Sensing (GIS/RS) Laboratory at ATREE support the
organization’s mission through research, service and outreach. Members of this lab work closely with
other scientists at ATREE to develop and facilitate research projects. The lab was also recently named
an International Resource Centre for IDRISI by Clark University, USA.
The main functions of the GIS/RS laboratory at ATREE are:
1) Conservation Planning
At ATREE, GIS and Remote Sensing play a key role in conservation planning, particularly for the assessment
and mapping of biodiversity over large areas, assessment of threats to various components of biodiversity
and the identification of priority areas. ATREE's efforts also include examination of the degree to which
the representative ecosystems or eco-regions are protected under the current PA network, identification of
areas that are of high conservation value but not included in the protected area network and development
of indicators for the loss of biodiversity. Additional aspects under consideration are the size,
patchiness and connectivity of protected areas, vulnerability of various species and ecosystems, and
policy and institutional mechanisms for conservation.
Mapping Biodiversity is an important aspect of conservation planning:
At ATREE satellite imagery based NDVI and NDVI derived data have been used to develop regression models to
predict species richness of beetles, birds, butterflies and trees in specific parts of the Western Ghats.
Mapping of biodiversity, threats to biodiversity and institutional/policy analyses is being done for the
entire Western Ghats region with detailed micro-level studies at three sites in the Western Ghats and
associated hills: Kalakad-Mudanthurai Tiger Reserve, Tamil Nadu and the Biligirirangan and Malai
Mahadeshwara hills in Karnataka. Tools to characterize landscape heterogeneity and vegetation classes
developed at ATREE will be used in these models.
Mean NDVI of Three Seasons
Using archival maps based on remotely sensed imagery and satellite data on forest cover, ATREE staff have
developed a detailed forest cover map of Arunachal Pradesh. Distribution of rare, endangered and endemic
species of orchids, rhododendrons, bamboos, birds and mammals has also been mapped. ATREE researchers
have modeled rates of deforestation and identified potential areas that need protection.
ATREEs mapping efforts have been at the local, regional and national level. Comprehensive economic
richness maps have been developed for Triphala (Phyllanthus emblica, Terminalia chebula and Terminalai
bellerica) and bamboos at BRT and for medicinal plants at Charmadi. We have initiated efforts in mapping
the biological resources and genetic variation of bamboo, rattan and Triphala at a regional level, i.e.
the central Western Ghats in south India. At a national level, we have been trying to map the spatial
distribution of bamboo, rattan and Phyllanthus and construct the contours of conservation.
2) Studying Land Use And Land Cover Change
The objectives of ATREE’s program on land use change are to assess the extent of change in land cover,
particularly deforestation and forest degradation, identify the causes of change, and examine their
consequences. It is concentrated in two regions: the Western Ghats and the Eastern Himalayas.
Using remote sensing imagery, archival maps, and ground surveys, scientists associated with ATREE have
examined land use change and deforestation in four areas: Agasthyamalai Hills, Malai Mahadeshwara Hills,
and Biligiri Rangan Hills in the Western Ghats, and Darjeeling Hills in the Eastern Himalayas. Detailed
maps showing the nature and extent of change have been prepared, and statistics on deforestation, forest
fragmentation, and conversion of forestland to other types of uses have been compiled. The data gathered
has been assembled into a Geographic Information System (GIS) format.
Working with the National Remote Sensing Agency (NRSA), Hyderabad, ATREE scientists have shown that the
rate of deforestation in the Western Ghats remains high. In Arunachal Pradesh, modeling has shown that
unless current trends are reversed, a large amount of biodiversity will be lost forever. ATREE has used
a method for identifying conservation-priority areas based on a predictive land-use change modeling
approach.
ATREE staff scientists have also examined deforestation in past and pre-independence periods from
historical records.Change in deforestation rates over time, particularly during the last few decades,
remains to be documented and is a high priority area for ATREE. Information about the extent of land
use change and causes and consequences of such changes are critical for inputs to new policies regarding
land use and conservation.
3) Research On Forests And Watershed Services
ATREE along with its partners Centre for Inter-Disciplinary Studies in Environment and Development (CISED),
UNESCO, National Institute of Hydrology and Karnataka Forest Department has initiated a
multi-institutional, multi-disciplinary and stake-holder-linked research project that will carefully
examine the poorly understood link between land-use/land-cover changes and watershed services at the
local and regional scale in the Western Ghats region. The study will be carried out in two sub-regions
of the Western Ghats that represent distinct combinations of rainfall regimes, forest types, soils,
land-use strategies, and the social importance of watershed services.
4) Development Of Tools & Approaches For Conservation
Classifying and mapping forests
ATREE has developed techniques of classifying forests in ways that take into account the fact that they
are not mosaics of discrete categories of vegetation types, but are a continuously changing terrain of
biological diversity. In addition, tropical forests with a distinct dry season have varying degrees of
deciduousness and density. The new method uses the mean and Coefficient of Variation (CV) of NDVI
across seasons. Areas of high NDVI and minimal CV can be considered to correspond to moist evergreen
forests while those with low mean and high CV would represent scrub forests. Thus the mean NDVI and CV
can be used as two different axes to represent a continuous gradient and once this method is standardized
it can be applied to areas where limited information on forest types is available. ATREE is also
experimenting with use of hybrid methods of vegetation mapping using combination of conventional
supervised classification of original four bands and rule-based classification based on two-date NDVI.
Mapping Forest Fires in Protected Areas
ATREE has successfully tested a methodological framework for post fire detection and mapping fire in the
different vegetation types of BRT WLS combining use of two-date remotely sensed NDVI data, a reference
eco- climatic/vegetation map for stratification and stratified random training data from a detailed ground
survey. ATREE generated fuzzy or probability maps using four different probability techniques: logistic
regression models, Generalized additive logistic regression model (GAM), CART tree-model and Bayesian
soft parametric likelihood algorithm. Probability values from all the four maps were broadly similar in
overall patterns and accuracy with respect to reference data but differed in some important respects.
The models generated can be used with pre and post fire season satellite imagery to map burnt areas
annually.
Quantifying landscape heterogeneity
ATREE is developing a technique to classify landscape heterogeneity using remotely sensed data. A grid
cell of 1 km2 that contains a greater number of vegetation classes that are eco-climatically very
different from each other is likely to provide habitat for a greater diversity of plants and animals.
Similarly a grid cell with different shapes and sizes of patches compared to a more homogeneous grid
cell is likely to provide habitats for larger number of species.
Landscape heterogeneity indices based on multi-season NDVI
ATREE is developing a technique to classify landscape heterogeneity using remotely sensed data. An index
developed at ATREE takes into account both these features as well as other measures of heterogeneity.
This index has been successfully tested for its ability to predict butterfly and tree diversity within
the BRT Wildlife Sanctuary in Karnataka. This technique will enable identification of localized hotspots
of biodiversity embedded within a given area.
ATREE’s ongoing research in these areas is expected to generate statistical models to predict various
components of biodiversity and a map of complex vegetation types for the entire Western Ghats. The index
can also be used to compare protected and unprotected areas.
|
