Below is a listing of projects that have utilized the Mapping Sciences Lab.

Funded Projects

NASA Forestry Project

The current and future trends in remote sensing present compelling opportunities for applying geo-spatial methods to forest resource problems. Read more about the NASA Forestry Project on their page.

NOAA - MERHAB

The National Oceanic and Atmospheric Administration (NOAA) through its National Center for Coastal Ocean Science (NCCOS) has developed the Monitoring and Event Response for Harmful Algal Blooms (MERHAB) project. One of SUNY ESF's contributions to this project is the development of procedures and techniques to efficiently detect and test for harmful algal blooms on the Great Lakes using remotely sensed images.

Recently Completed Research

McIntire-Stennis RADAR

Satellite RADAR data is becoming more prolific and less expensive. The objective of this project is to examine the uses of RADAR for forestry classification applications to take advantage of the data availability. The researchers are currently using data from ERS and JERS missions to discriminate hardwood, softwood and mixed wood forest.

NASA Affiliated Research Center (ARC)

Funding for the NASA ARC ceased on 12/31/2002 however, the work performed by the Affiliated Research Center at SUNY-ESF is still on the cutting edge of remote sensing/GIS research. Navigate to their site to get a detailed description of the work they did.

George Washburn

In conjunction with the Canal Design Bureau of the New York State Thruway Authority and the New York State Department of Transportation, George explored the viability of using multispectral imagery from a Daedalus Airborne Multispectral Scanner and automated image classification techniques to assess the condition of the embankments on the NYS Barge Canal system.

Jennifer Barber

Using 1 meter resolution CIR imagery from Emerge, Jennifer examined the practicality of applying two-dimensional Fourier transformations to classify broad leaf and needle leaf forest canopy types. The research was performed on data acquired over the Huntington Memorial Forest in the Adirondack Park.

Miguel Reynero

Landsat TM5 images of central Argentina were used by Miguel to determine biomass loss and recovery capacity of burned grasslands. An NDVI and Infrared Index were applied to a set of six TM5 images spanning ten years.

Elizabeth Coyle

Starting with soil maps, terrain models, hydrologic features, and storm flow measurements, Liz developed a model for measuring non-point source pollution entering the Onondaga Lake watershed using commercial GIS packages.

Marian Poczobutt

Improving land cover classification in areas of varying slope and aspect was the goal of Marian's research. Using Landsat TM images and digital elevation models, Marian stratified the images and then compared the land cover classification accuracy of the strata.

Kerry Van Siclen-Turick

Kerry, while a graduate student within the Faculty of Environmental Resources and Forest Engineering at SUNY ESF, did extensive work at SUNY Upstate Medical University in Syracuse. There, she assisted MD's and other medical researchers develop tools to systematically detect breast tumors using data from Magnetic Resonance Imaging (MRI). The ultimate goal of this and continuing research is to find a non-invasive technique that can consistently differentiate between malignant and benign tumors.

Lester Power

Coniferous forests in North Carolina and Oregon were the sites of test ranges where Lester performed a study of GPS receiver accuracy and efficiency under the conifer canopy. Three techniques, moving the antenna, reducing the signal-to-noise ratio mask and using an amplified antenna, were investigated. Lester found that moving the antenna actually improved point accuracy, and the other methods, while showing no substantial increase or decrease in accuracy, increased the efficiency in which point data could be collected.

Lindi Quackenbush

Lindi, completed here PhD. in the spring of 2004. Her research focus was on delineating subclasses of impervious land cover surfaces in remotely sensed data using a technique known as Fractal Dimension. This technique has been useful in measuring complex coastlines and terrain surfaces. It is expected that fractal dimension will also help characterize the differences between the various subclasses of impervious surfaces.

William Stiteler IV

Bill, developed genetic algorithms and cellular automata to aid in image processing and landcover classification. Bill has developed these algorithms to mimic natural evolution. Using an unsupervised classification of an image as the base, multiple iterations of the algorithm transform the raster cell values to a point where they closely match ground referenced data. Bill earned his PhD. in December of 2003.

Cheng Zhu

Data fusion is the topic of Cheng's research for her PhD. She is investigating the advantages of fusing optical satellite imagery with satellite based synthetic aperture radar (SAR) in forest classifications. Using artificial neural networks with JERS-1 L-band SAR data and Landsat7 ETM images, Cheng hopes to more accurately distinguish hardwood and softwood classes from mixed classes.

Ongoing Research

Mark Pugh

Using a variety of commercial software packages, Mark, a PhD. candidate, is developing techniques for feature identification and terrain classification using fuzzy neural networks with multi-sensor image fusion. By properly combining information from different sensor modalities, Mark believes it should be possible to improve the accuracy and robustness of landcover classification.

Pauline Stephens

Pauline, an MS candidate, is concentrating her efforts on the MERHAB project. At this time she is performing an assessment of aerial/satellite sensor capabilities with respect to identifying algal blooms in the Great Lakes.