Mapping The Extent and Distribution of Palmetto Palm in Three Protected Areas of Southern Belize Using Remote Sensing Techniques

Abstract

Across Belize, organisations are working towards effective management and control of protected conservation areas. There is a drive to look after and enhance biodiversity and cultural heritage, whilst also managing forest and other resources in a sustainable way. To effectively plan and manage this, organisations need to know what different land cover types and resources they are working with, and therefore need accurate up-to-date mapping of each area. The Toledo Institute for Environment and Development work alongside the Government of Belize Forest Department to manage a region in the district of Toledo in Southern Belize, which contains three protected areas (PAs) of varying size and terrain. This study has identified the location and extent of palmetto palm, one of the key resources within the PAs, using remote sensing techniques. This is important as it will help with a number of different management plans such as identifying palmetto areas suitable for harvesting by local communities. A maximum likelihood classifier was used on WorldView-2 multispectral imagery to identify the key vegetation types including the palmetto in the three areas. A number of problems were identified with this process and the results produced, through comparison with ground data. Due to the overlapping of spectral signatures between the main vegetation and land cover types, large areas were wrongly identified in the classification process. A number of improvement methods were explored, which reduced the extent of the areas that had mis-classified as palmetto. This reduced overestimation of the resource and increased the user accuracy for palmetto, however there were still areas of confusion occurring between the main vegetation types. Further research needs to be done to solve this problem. Maps were produced with the final palmetto areas identified, which had a user accuracy of over 80%.