Vegetation and GIS Land Use/Land Cover Analysis
based upon Satellite Remote Sensing and Extensive Field Survey
Robert Zomer
Ph.D.Candidate
Ecology Graduate Group
University of California
Davis, CA, 95616, USA
rjzomer@ucdavis.edu
-------------------------
Graduate Advisors:
Dr. Jack Ives
Division of Environmental Science
University of California, Davis, CA, 95616, USA
Dr. Susan Ustin
Center for Spatial
Technologies and Remote Sensing
Dept. of Land, Air and Water Resources
University of California, Davis, CA, 95616, USA
Dr. John Menke
Dept. of Agronomy and Range Science
University of California, Davis, CA, 95616, USA
Click here to see more images...
Abstract
The Makalu-Barun National Park and Conservation Area (MBNPCA) was established
in 1992 to preserve unique biodiversity and endangered wilderness in Nepal,
and the Himalaya. This unique mountainous area of over 2300 sq. km comprises
the drainage extending from the Mt. Everest ecosystem (elev. 8000 m) to
the Arun River Valley (elev. 300 m), encompassing an enormous diversity
of species, habitats, and bioclimatic regions. The park management plan
calls for sustainable development for the more than 32,000 agriculturalists
living within the Conservation Area. Like many biosphere reserves in developing
countries, managers lack reliable information. Remote sensing satellite
imagery will be utilized to develop a Geographic Information System (GIS)
for the MBNPCA. Landsat TM and SPOT digital imagery, and extensive vegetation
and landuse field survey data collected on six research expeditions (1991-1994),
will be used to classify and delineate vegetation-types, land-use, and
wildlife habitat within the MBNPCA . A comparative baseline and initial
information database will be developed.
Objectives
A Geographic Information System (GIS) for the Makalu-Barun National Park
and Conservation Area (MBNPCA) based on land-cover type classes will be
developed. Vegetation communities and forest structure will be surveyed,
classified, and analyzed based upon an extensive set of georeferenced fieldwork
data collected on six previous expedition to the MBNPCA. Stereoscopic SPOT
satellite remote sensing data will be used to extract a digital elevation
model (DEM), and a terrain corrected base map layer developed for the MBNPCA
(accurate topographic map are currently not available). Landsat TM satellite
multispectral data will be analyzed to determine vegetation/cover type,
and incorporated into a GIS environment (ARC/INFO), along with vegetation
and landuse field data. Other relevant data sets, as available, will be
incorporated into the GIS environment. This augmented satellite remote
sensing-based GIS will be used to:
Background:
The design and management of national parks requires that policy makers
and managers have access to both a broad base of information, and adequate
analytical tools (Yonzon et al., 1991). In many parks, reliable information
on park physical attributes, extent and type of vegetation , and land use
patterns is lacking, due to physical remoteness and/or lack of funds and
personnel for research. In this respect, Makalu-Barun National Park is
typical of biosphere reserves in impoverished developing countries. By
incorporating several recent technical advances with extensive field work
and satellite remote sensing data, a cost-effective resource management
and analysis methodology is proposed.
Remote sensing provides an efficient and cost-effective method of acquiring
up to date and accurate landscape and regional level information for use
by resource planners. Global Positioning System (GPS) technology now offers
an efficient and cost-effective ground truthing and land survey ability.
In order to effectively utilize such information, however, resource managers
need tools capable of quickly and efficiently updating and analyzing spatial
data. GIS technology is a spatial database approach designed to meet these
needs. Incorporating university-level field-research courses into GIS field
data collection efforts is a cost-effective solution to acquiring the labor
intensive field data required to develop an accurate GIS natural resource
database, while providing unique field-research opportunities to college
under-graduates.
Study Area
The Makalu-Barun National Park (est. 1992) and Conservation Area (MBNPCA)
of Nepal , located on the eastern slope of the Mt. Everest ecosytem, is
the most recent major conservation effort aimed at preserving biodiversity
and threatened wilderness within the Himalaya. Habitats ranging from tropical
monsoonal rain forests to alpine tundra and peaks reaching over 8000 m
are located in close proximity . Over 3,000 species of flowering plants,
including 25 of Nepal's 30 varieties of rhododendron, 48 primroses, 47
orchids, 19 bamboos, 15 oaks, 86 fodder trees and 67 economically valuable
aromatic and medicinal plants have been reported (Polunin and Stainton,
1984; Stainton, 1988; Shrestha, 1989). Several species of endangered wildlife
are found in the park as well, including red panda, musk deer, barking
deer, ghoral, flying squirrel, thar, common leopard, and snow leopard.
Bird fauna is rich and diverse, ranging from a wide diveristy of eagles
and other raptors to white-necked storks and brilliantly colored sunbirds.
Several species of endemic fish have been reported as well. The Barun Valley
has been recognized for itÕs geographical uniqueness and species
rich biodiversity, and classified a UNESCO World Heritage Site. Several
investigators have categorized vegetation within the eastern Himalaya (Stainton,
1972; Dobremez, 1976; Numata, 1983, Schweinfurth, 1984; Shrestha, 1989;
Schmidt-Vogt, 1990; Miehe, 1991). In addition to the high level of habitat
and species diversity, the park also encompasses considerable cultural
diversity and richness. More than 32,000 people live within the Conservation
Area (approximately 830 sq. km), representing several distinct ethnic groups
and languages. The vast majority are subsistence agriculturalists, either
sedentary, swidden, or trans-humance pastoralists. Local peoples depend
heavily upon adjacent natural resources (Ives and Messerli, 1989; Gilmour
and Fisher, 1991; Brandon and Wells, 1992), including grazing and animal
fodder collection, fuelwood and timber collection, and collection of medicinal
and other economically valuable plants.
SFSU Wildlands Studies - Nepal Program:
The Wildlands Studies - Nepal Program (San Francisco State University Extension)
serves as the primary focus for field work and ground-truthing activities
detailed in this proposal. The Wildlands Studies Program conducts a university-accredited
Environmental Field Studies course (nine semester-units) in the Makalu-Barun
National Park. Incorporating this course into a larger research framework
has both provided meaningful context to this unique field studies course,
and offered opportunity for extended field study and intensive data collection.
Robert Zomer (graduate student named in this proposal) is an instructor
and coordinator for the program. To date, six successful courses have been
completed, for a total of moe than 240 field days by 12 to 16 American
college students and staff, along with substantial local support staff
(including local experts, four local botany grad students (M.Sc.) from
Tribhuvan University in Kathmandu, and porters and guides). The Wildlands
Studies courses are self-supporting and cover most of the fieldwork expenses.
Methodology
Field Studies
During six previous research expeditions to the MBNPCA: Preliminary cover
classifications were determined through vegetation and landuse field surveys.
Extensive field data was collected on 256 representative forest plots,
as well as several dozen transects in alpine meadows. Vegetation patterns
were recorded at stratified random field locations, using a methodology
partly based upon, and congruent with, Shrestha et al. (1990) and his earlier
work in the MBCP. Plots of 400 m2 were inventoried for all woody vegetation
with stems greater 10 cm dbh. Percent cover of both the understory and
canopy was estimated and characterized. Site parameters (elevation, slope,
aspect, terrain) were determined. The nature and extent of landuse activities
was categorized based upon observation and informal interviews. Global
Positioning Receivers, with external antennas on 10 m collapsible poles,
were used to estimate geographic coordinates for the plot. This geo-referenced
plot data will be utilized as the training set in a supervised multispectral
classification for the remote sensing data.
Community Analysis
Multivariate clustering techniques (Gauch, 1982), two-way indicator species
analysis (TWINSPAN), and canonical correspondence analysis (CANOCO) (Jongman
et al., 1987) will be used to analyze community ecology, and determine,
describe, and delineate vegetation classes.
Image processing and analysis
Stereoscopic SPOT MSS data will be used to extract a DEM and develop a
terrain corrected digital base map, using PCI Satellite Orthorectification
software. Landsat TM digital data (Sept. 1992) will be used as the primary
data set for vegetation mapping. Other remote sensing data sets include
Landsat MSS data from 1979, to add a temporal dimension to the analysis,
and SPOT MSS data for added spatial resolution and to aid in identification
of agricultural and settlement patterns.
The ability to identify and evaluate landscape-level spatial and temporal
processes has been greatly enhanced with the advent of satellite remote
sensing. Satellite remote sensing has been used to map and classify landuse
and vegetation cover classes in remote, inaccessible areas and mountainous
regions (Frank, 1988; Lal et al., 1991), including Nepal (Blamont and M*ring,
1987). Inclusion of topographic data, has been shown to improve prediction
accuracy significantly (Strahler et al., 1978, Talbot and Markon, 1986;
Frank, 1988; Senoo et al. 1990). By incorporating the remote sensing data
into a GIS before classification, other ancillary data sets may be utilized
in a decision-making tree, or rule-induction approach to classification.
This method was found useful in disturbed and hilly environments (Lees
and Ritman, 1991).
GIS development
All remote sensing data will be incorporated as distinct layers within
the GIS environment (ARC/INFO), and merged with GIS base map and ancillary
data sets provided by ICIMOD (MENRIS). The extracted DEM will be used to
aid in classification of land cover types. Field study data will utilized
as a training set to produce a supervised multispectral classification
based on all relevant data sets within the GIS. Ancillary data sets, as
available, including census data, administrative boundaries, and socioeconomic
data, will also be incorporated as distinct layers or georeferenced database
information.
Modeling and analysis of willdlife habitat
Delineate and quantify potential wildlife habitat, habitat quality, and
habitat fragmentation, for several species of charismatic megafauna (e.g.
red panda, snow leopard, musk deer), using a GIS modeling of known wildlife
habitat relationships. Potential and actual habitat has been modeled within
a GIS environment for a diverse range of wildlife species (Tomlin et al.,
1987; Young et al., 1987; Hodgsen et al. 1988; Agee et al., 1989, Prasad
et al, 1992; Yonzon et al, 1991).
References
Agee, J.K. Stitt, S., Nyquist, M and Root, P. 1989. A geographic
analysis of historical grizzly bear sightings in the North Cascades. Photogram
Eng. Remote Sens. 54: 1601-1607.
Blamont, D. and C. M*ring, 1987: Use of remote sensing
for vegetation and landuse mapping in mountainous areas: The case of central
Nepal. Adv. Space Res. 7(3): 41-46.
Brandon, K.E. and M. Wells. 1992: Planning for people
and parks: design dilemmas.. World Development. 20 (4): 557-570.
Carpenter, C. and R. Zomer, 1996. Forest Ecology of the
Makalu-Barun National Park and Conservation Area, Nepal. Mountain Research
and Development 16 (2): 135-148.
Dobremez, J.F. 1976: Le Nepal: ecologie et biog*ographie.
Cahiers N*palais. Paris.
Frank, T.D. 1988: Mapping Dominant Vegetation Communities
in the Colorado Rocky Mountain Front Range with Landsat Thematic Mapper
and Digital Terrain Data. Photogrammetic Engineering and Remote Sensing,
54(12): 1727-1734.
Gauch, Jr, H.G., 1982: Multivariate analysis in community
ecology. Cambridge University Press, 298 pp.
Gilmour, D.A. and R.J. Fisher, 1991: Villagers, Forests
and Forests: The philosophy, process and practice of community forestry
in Nepal. Sahayogi Press. Kathmandu, Nepal. pp. 212.
Hendrix, W.G., Fabos, J.G. and J.E. Price, 1988: An ecological
approach to landscape planning using geographic information system technology.
Landscape and Urban Planning, 15:211-225.
Hodgson, M., Jensen, J., Mackey, H., and Coulter, M. 1988.
Monitoring wood stork foraging habitat using remote sensing and geographic
information systems. Photogram Eng. Remote Sens. 55: 1637-1642.
Ives. J. D. and B. Messerli. 1989: The Himalayan Dilemma:
Reconciling development and conservation. The United Nations University.
Routledge, New York.
Jongman, R.H.G., ter Braak, C.J.F., and O.F.R. van Tongeren,
1987: Data analysis in community and landscape ecology. Pucdoc, Wageningen.
299 pp.
Lal, J.B., Gulati, A.K. and M.S. Bist, 1991: Satellite
mapping of alpine pastures in the Himalayas. Int. J Remote Sensing. 12
(3): 435-443.
Langley, P.G. 1971. Earth resources information systems
using satellites, aircraft and ground teams. Proc. of the Twenty-first
Int. Astronautical Congress. North-Holland.
Lees, B.G. and K. Ritman, 1991: Decision-tree and rule-induction
approach to integration of remotely sensed and GIS data in mapping vegetation
in disturbed or hilly environments. Environ. Management 15(6): 823-831.
Miehe, G., 1991: The vegetation map of the Khumbu Himal
(Mount Everest south slope) 1:50 000. Vegetation patterns and problems
of mapping. Gefugemuster der Vegetation und Probleme der Kartierung. Erdkunde,
45(2), pp 81-94.
Numata, M. 1983: Structure and dynamics of vegetation
in eastern Nepal. Laboratory of Ecology, Faculty of Science, Chiba University,
Japan. 184 pp.
Polunin, O. and A. Stainton, 1984: Flowers of the Himalaya.
Oxford University Press, Delhi, India pp. 580.
Prasad, S. N., Chundawat,R. S., Hunter, D. O, Panwar,
H. S., Rawat, G. S., and G. J. Buhyoff, 1991: Remote sensing snow leopard
habitat in the trans-Himalaya of India using spatial models and satellite
imagery preliminary results. In: Resource Technology 90. Proc. second international
symposium on advanced technology in natural resources management, Washington,
DC, 1990. pp . 519-523.
Senoo, T., Kobayashi, F., Tanaka, S., and T. Sugimura..
1990: Improvement of forest type classification by SPOT HRV with 20 m mesh
DTM. Int. J. Remote Sensing 11(6): 1011-1022.
Shrestha, T.B. 1989: Development Ecology of the Arun River
Basin in Nepal. International Centre For Integrated Mountain Development,
Kathmandu, Nepal.
Shrestha, T.B., R. Sakya and H.S. Nepali, 1990: Scientific
Report on field survey of 1989: General and Phyto-ecology Working paper
No. 8, Makalu-Barun Conservation Project, Kathmandu, Nepal.
Stainton, A., 1988: Flowers of the Himalaya: A Supplement.
Oxford University Press, Delhi, India. pp. 85.
Stainton, J.D.A. 1972: Forests of Nepal. J. Murray Ltd.
London. 181 pp.
Strahler, A., Logan, T.L., and Bryant, N.A., 1978: Improving
forest classification accuracy from Landsat by incorporating topographic
data. In; Proc. Twelfth Int. Symposium on Remote Sensing of Environment.
Ann Arbor, Mich., p. 927-942.
Talbot, S.S. and Markon, C.J. 1986: Vegetation mapping
of Nowitna National Wildlife Refuge, Alaska using Landsat MSS Digital Data.
Photogrammetic Engineering and Remote Sensing, 52 (6): 791-799.
Tomlin, S.D., Berwick, S.H. and Tomlin, S.M. 1987. The
use of computer pgaphics in deer habitat evaluation. In: Geographic Information
Systems for Resource Management: A Compendium. Ripple W. (ed.) American
Society for Photogrammetry and Remote Sensing and American Congress on
Surveying and Mapping, Falls Church, Va., p.145-156.
Yonzon, P., Jones, R., and J. Fox, 1991: Geographic Information
Systems for Assessing Habitat and Estimating Population of Red Pandas in
Langtang National Park, Nepal. Ambio. 20(7): 285-288.
Young, T.N. , Elby, J.R., Allen, H.L. Hewitt, M.J. and
Dixooooon, K.R. 1987. Wildlife habitat analysis using Landsat and radiotelemetry
in GIS with applicaton to spotted owl preference for old growth. In: Proc.
of GIS '87 San Francisco. American Society for Photogrammetry and Remote
Sensing. Univ. of Calif. Berkeley, p. 595-600.
Zomer, R. and J. Menke, 1993. Site Index and Biomass Productivity
Estimates for Himalayan Alder - Large Cardamom Plantations: A Model Agroforestry
System of the Eastern Mid-Hills of Nepal. Mountain Research and Development
13 (3): 235-255.
Back to Landscape Ecology of Makalu-Barun