Date of Award
Master of Science in GIS for Development and Environment (GISDE)
International Development, Community and Environment
Hurricane damage to tropical forests has immediate impacts such as branch and trunk damage or tree uprooting, and long-term impacts including mortality, and changes in species composition. Satellite imagery can provide regional-scale, continuous coverage of hurricane damage to locate areas of high impact and track forest recovery. This study identifies immediate disturbance to forest types from Hurricane Dean, category 5 on the Saffir-Simpson scale, in 2007 in the southern Yucat´an Peninsula. Spectral mixture analysis applied to 500 m 8-day surface reflectance imagery from the Moderate Resolution Imaging Spectroradiometer (MODIS) is used to map changes in fractions of green vegetation (GV) and non-photosynthetic vegetation (NPV) throughout the hurricane impacted area. GV loss and NPV gain varied by wind speed and forest type. Tall and medium forests covered the largest area that experienced coincident high GV loss and NPV gain. Throughout all forest types, the highest mean GV loss occurred in modeled wind speeds of 60-62 m/s. The highest mean NPV gain throughout all forests occurred in wind speeds of 75-77 m/s. Forests experienced a higher mean GV loss north of the track at a distance of up to 10 km from the track and 25-80 km from the track. At 25-80 km from the hurricane track, forests experienced a higher mean NPV gain north of the track as opposed to south of the track. The results of the study indicate that locating hurricane damage using satellite imagery might benefit from using multiple indicators that are sensitive to both the decrease in green vegetation and the increase in woody debris.
Khan, Anam, "Identifying factors in differential hurricane damage to the forests in the Southern Yucatan Peninsula using MODIS imagery" (2017). International Development, Community and Environment (IDCE). 202.