Alumna Profile: Meghan Graham MacLean
Degree: Natural Resources and Environmental Studies
Dissertation Title: A Multi-Temporal Image Analysis of Habitat Modification in the Coastal Watershed, NH
Advisor: Dr. Russell Congalton
Description of Research
As the human population has grown significantly in the last 20 years throughout New England, Southeastern New Hampshire has become increasingly susceptible to issues associated with urban development and sprawl. Urbanization has increased the amount of impervious surface in the area as well as fragmented many of the crucial habitats, especially forests. The fragmentation of these habitats can have negative effects on both the vegetation and the wildlife in that area, as well as human quality of life. Therefore, mapping and analyzing these fragmented habitats is of critical importance for conserving and protecting critical habitat for wildlife, water quality, and many other natural resources. Given the recent no-cost availability of Landsat data, a multi-temporal analysis of changes in habitat over time is now much easier. My project proposes to quantify land cover change and forest fragmentation in the Coastal Watershed of NH using Landsat data of this area from the last 25 years.
Current land cover data of the area either does not have map the area in enough detail to determine how specific habitats are reacting urbanization, or do not have enough temporal resolution. Therefore, my first objective of my research will be to create new land cover maps from Landsat images using an object-based image analysis approach to start quantifying land cover change in the Coastal Watershed. The maps will have specific land cover classes, including forest categories defined by dominant tree species. Sample sites are currently being collected both on the ground as well as through photo interpretation. These samples sites will be used for both training data and accuracy assessment sites. New land cover maps will be created from Landsat images for every three years from 1985 until 2010 to get a historical perspective of land cover change in this area.
The next objective of my research will be to create a new fragmentation program to be implemented in ArcGIS 10. The new program will combine the strengths of some current fragmentation tools, but increase the usability of these types of programs. The new tool will output a map of forest fragmentation, but also allow the researcher to quantify how each individual habitat class is changing or being fragmented over time. This will allow me to discover whether the urbanization of the Coastal Watershed is affecting specific land cover types more than others and possibly correlate the current location of invasive species within the watershed with past fragmentation of certain habitats. Lastly, the record of fragmentation in the area can be used to make future predictions about the changing landscape given certain parameters for climate change, urbanization, etc. My hope with my dissertation is to finish with a product that will be useful not only to researchers, but also to policy makers, conservationists, and land use planners as they make decisions about our landscape.
Previous Awards and Activities
- Chair of the American Society for Remote Sensing and Photogrammetry (ASPRS) Student Advisory Council, 2010 – 2011
- Member of ASPRS and AAG
- University of New Hampshire Graduate School Travel Award, 2008 – 2010
- University of New Hampshire Graduate School Summer Teaching Assistant Fellowship, 2008 – 2010
Farrington Fund Travel Grant, 2008 & 2009
Selected Presentations and Publications
- Graham, M.E., 2008. Evaluating accuracy issues in mapping benthic habitats: an investigation in the causes of misclassification and the importance of segmentation parameters. MS Thesis, University of New Hampshire, Durham, NH. 113 p.
- Graham, M. and R. Congalton, 2009a. Evaluating issues in map accuracy: A study of mapping benthic habitats on the Texas gulf coast. In: ASPRS 2009 Annual Conference, 9-13 March 2009, Baltimore, MD (Bethesda: American Society for Photogrammetry and Remote Sensing), pp. 23-34.
- Graham, M. and R. Congalton, 2009b. A comparison of the 1992 and 2001 National Land Cover Datasets in the Lamprey River Watershed, NH. In: ASPRS/MAPPS 2009 Fall Conference, 16-19 November 2009, San Antonio, TX (Bethesda: American Society for Photogrammetry and Remote Sensing).
- MacLean, M.G., A. Rudko, and R. Congalton, 2010. Multi-temporal image analysis of the Coastal Watershed, NH. In: ASPRS 2010 Annual Conference, 26-30 April 2010, San Diego, CA (Bethesda: American Society for Photogrammetry and Remote Sensing).
- MacLean, M.G., and R. Congalton, 2010. Mapping and analysis of fragmentation in southeastern New Hampshire. In: ASPRS/CaGIS 2010 Fall Specialty Conference, 15-19 November 2010, Orlando, FL (Bethesda: American Society for Photogrammetry and Remote Sensing).
- MacLean,M.G., and R. Congalton, 2011. Using object-oriented classification to map forest community types. In: ASPRS 2011 Annual Conference, 1-5 May 2011, Milwaukee, WI (accepted).
Ph.D. University of New Hampshire ’12 - Natural Resources and Environmental Studies
M.S. University of New Hampshire ’08 – Natural Resources and the Environment
B.S. Clarkson University ’06 – Civil & Environmental Engineering and Physics