PARENT SESSION
Oral Session # 72: GIS and Remote Sensing I.
Presiding: J Drake
Thursday, August 7. 8:00 AM to 11:30 AM, SITCC Meeting Room 203.

Application of a Pseudo-NDVI image enhancement to coastal landscape ecology.

McMillan, Brett¹, Bellows, A. Scott¹, Allen, Thomas², ¹ Biological Sciences, Norfolk, VA² Political Science and Geography, Norfolk, VA

ABSTRACT- The Normalized Difference Vegetation Index (NDVI) is a long and widely-used vegetation index applied to combinations of spectral bands from digital remotely-sensed images and used to accentuate and measure vegetation characteristics. NDVI uses red and infrared wavelengths, such as the multi-spectral sensors on Landsat, to detect chlorophyll and leaf structure variations in vegetation. The coarse spatial resolution of most satellite observations (e.g., 30 x 30 m pixels from Landsat Thematic Mapper) makes them most appropriate for landscape-level investigations on a regional scale. Although multi-spectral satellites with finer spatial resolutions are becoming available, they often lack sensing abilities in the infrared wavelengths that are relevant to plant ecology. The primary historical source of multi-spectral imagery with pixel size < 5 x 5 m is color-infrared (CIR) aerial photographs, which lack visible red wavelengths. Having a need to detect changes in vegetation over short distances (< 2 m) on a small island, we developed a modified or pseudo-NDVI (P-NDVI) for use with CIR aerial photographs. Our P-NDVI substitutes blue band values for red band values in the same function used for original Landsat NDVI, with the rationale that chlorophyll absorbs blue and red wavelengths nearly equally. Furthermore, blue light scattering, a common problem with satellite images, is reduced in aerial photography because of lower observational altitude. Classification of vegetation at test locations using NDVI on Landsat images and P-NDVI on aerial photographs produced comparable results. This protocol has an advantage over multivariate statistical classification procedures such as principal components analysis (PCA), in that detected variation can be attributed to true differences in vegetation with greater confidence. This technique should be useful for conducting coastal vegetation and plant ecology studies at a finer spatial scale and over a longer time range than is available with satellite imagery.

Key words: Coastal Ecology, Remote Sensing, NDVI, Plant Ecology