Jennifer Miller, who earned her PhD through the joint program between San Diego State University and UC Santa Barbara in 2003, wrote a paper, based on her dissertation, that won a Nystrom Award at the Association of American Geographer’s (AAG’s) conference in March 2004.* She co-wrote the paper with Janet Franklin. It is titled “Incorporating spatial dependence in predictive vegetation models: residual interpolation methods.”
Miller’s paper is about researching a potential way to increase the accuracy of predicting what kind of vegetation might be where and how abundant it might be. Because of the high cost of mapping in the field, scientists and policy-makers want to find models that can accurately predict the vegetation, using remote sensing and GIS data about elevation, precipitation, temperature, slope, and landforms, etc. In order to be good stewards of natural resources, not only do we need to know what is there, but we need to be able to predict what might be present with climate changes, including global warming.
Miller researched the possible benefits of incorporating residuals into the vegetation prediction model, following methods that have been used in soil prediction modeling. The results, according to Miller, “were somewhat mixed, mostly as a result of the varying proportions of the vegetation types studied. I looked at 11 types, ranging from very rare (less than 1% of the sample) to common (approximately 40% of the sample). The rare vegetation types often did not have enough data to support … interpolation.”
Miller started the UCSB/SDSU joint program in 1997 and finished summer 2003. At UCSB, she worked with Michael Goodchild and Joel Michaelsen. Her dissertation had basically the same title as the paper that won the award: “Incorporating Spatial Dependence in Predictive Vegetation Models.” As a geographer, her interests are GIS, spatial analysis, and biogeography. She began an assistant professor position at West Virginia University in GIScience starting fall 2003.
Miller is also received an award for “Best student presentation” at the March 30 – April 4, 2004, International Association of Landscape Ecology (U.S. Chapter) conference in Las Vegas. The award is for her presentation at last year’s conference of her paper, titled “A comparison of methods for incoroporating spatial dependence in predictive vegetation models.”