Projects & Publications
Determine the Effectiveness of Automated Plant Community Detection From High Resolution Satellite Imagery
This research project supports the assessment of landscape patterns in the Greater Everglades wetland domain. RECOVER evaluates assessments of vegetation community structures and landscape patterns via various means: ground referencing and related mapping of field morphometrics, community and species identification along elevation gradients (marl prairie to ridge and slough gradients), and community typing via photogrammetry. The project’s scope focuses on the evaluation of methods developed using new satellite imagery (medium spectral and high spatial resolution). Methods and imagery evaluated include supervised classification algorithms, with the use of reflectance and texture variables derived from Digital Globe WorldView 2 imagery captured during both the wet and dry seasons. The project’s goal is to map plant communities at a 20-to 40-square-meter spatial resolution. The project’s objectives are to investigate and characterize the spectral and metric qualities of the WorldView 2 data as related to specific aspects of the sensor, data acquisition and level of processing, and the effects of spatial re-sampling on the spectral integrity of the imagery and on the type of landscape information that can be derived at various spatial scales. The results can then be compared to the results of processing medium to coarse resolution satellite imagery (e.g., Landsat 30 m).
Information gained from this project will provide future guidance relevant to the Comprehensive Everglades Restoration Plan (CERP) structure and implementation. The scope is written as part of the RECOVER Monitoring and Assessment Plan (MAP), Sections 3.1.3 and 3.1.4.
The results of the work performed will be essential to the refinement of existing Greater Everglades project performance measures, as well as linking the existing predictive metrics to the CERP Monitoring and Assessment Plan (RECOVER 2004, 2006b, 2008) and its associated hypothesis clusters.
- Years
- 2011
- Principal Investigator
- Daniel Gann
- Agencies
- South Florida Water Management District