Estimation of Haitian forest cover from 1995-2016 using supervised classification of Landsat data

Authors: Patrick Taylor*, Grand Valley State University
Topics: Remote Sensing, Environmental Science, Geographic Information Science and Systems
Keywords: Haiti, Forest Cover, Remote Sensing, ATCOR, Supervised Classification
Session Type: Poster
Day: 4/11/2018
Start / End Time: 8:00 AM / 9:40 AM
Room: Napoleon Foyer/Common St. Corridor, Sheraton, 3rd Floor
Presentation File: No File Uploaded

The purpose of this study is to show the change in the forest cover of Haiti in order to gain a more accurate representation on how its natural resources are being utilized. The World Bank currently has Haiti's forest cover at 4%, this is a misrepresentation of the actual forest cover in Haiti. This study uses Landsat images from 2010-2011, 1995-1996, and 2015-2016 to estimate the total forested area in Haiti. The Landsat images were first atmospherically corrected using Atmospheric and Topographic Correction (ATCOR) module. A thematic map was then generated by a supervised classification of the mosaicked image using the Food and Agriculture Organization (FAO) of the United Nations (UN) Land Cover Classification System. Classification results were ground truthed through comparison with multiple datasets . Based on our classification results, approximately 17% of Haiti’s total land area was tree covered in 2010-2011. This result was greater than the reported World Banks value of 4% and was less than the study conducted by Churches, C., Wampler, P. J., Sun, W., and Smith, A. J., 2014. Evaluation of forest cover estimates for Haiti using supervised classification of Landsat data. Our study suggests that at the coarse resolution of data used, the patchy and fragmented forest cover has possibly caused a systematic underestimation of the extent of forest cover.

Abstract Information

This abstract is already part of a session. View the session here.

To access contact information login