| This research builds upon years of investigation in the application of satellite remote sensing to wetland biomass estimation. While previous work focused on the quantitative estimation of standing biomass, image acquisition issues and natural environmental variability limited the utility of this type of information for change detection. In order to address this problem, this research describes the development of a spatial and temporal statistical technique to generate Locally Normalized Biomass Indicator (LNBI) values suitable to trends analysis.; The wetland LNBI trends methodology allows for the identification and monitoring of both long-term and event driven wetland perturbations. The developed model has several advantages including insensitivity to between scene variations due to time of acquisition and ground target conditions, as well as being able to identify changes within wetlands which may precede any change of dominant species composition. |
