Determining Species Composition Using Temporal NDVI Trajectories Derived from Satellite Remote Sensing Measurements
Abstract
Significant distinctions in phenological properties exist between plant communities. These differing characteristics can be the result of varying plant types, environmental stressors, or both of these factors. This paper explores the ability of satellite remote sensing to determine relative amounts of C3 and C4 vegetative lifeforms contained within a heterogeneous canopy based on their unique phenological qualities. Changes in the structure and dynamics of an environment, in this case a tallgrass prairie, can be indicators of even larger stresses on the ecosystem. The C3 and C4 plants of a tallgrass prairie, for example, have been shown to be particularly sensitive to environmental changes. The ability to distinguish between these life forms can be complex due to the fact that they appear spectrally similar at a single point in time. Other studies have proven this to be possible using hand held, close range remote sensing measurements with fine spatial resolution. However, the ability to distinguish these life forms using imagery of relatively coarse resolution has not yet been explored. Twenty-six SPOT images from March to November of 1998 were obtained for analysis. Percentages of C3 and C4 plants were determined using discriminant function mixture models based on metrics derived from the temporal trajectory of the Normalized Difference Vegetation Index (NDVI) . The percentages derived from satellite measurements were compared with data collected on the ground. Classification accuracies between 50%-60% were obtained using the techniques discussed in the paper. The results presented in this report seem to be a promising indicator that determining amounts of C3 and C4 within a canopy is possible using satellite remote sensing.
Document Details
- Document Type
- Technical Report
- Publication Date
- Aug 04, 1998
- Accession Number
- ADA350852
Entities
People
- Matthew J. Tracy
Organizations
- Kansas State University