Login / Signup

Estimating Structural Damage to Mangrove Forests Using Airborne Lidar Imagery: Case Study of Damage Induced by the 2017 Hurricane Irma to Mangroves in the Florida Everglades, USA.

Selena ChavezShimon WdowinskiDavid LagomasinoEdward Castañeda-MoyaTemilola FatoyinboRyan P MoyerJoseph M Smoak
Published in: Sensors (Basel, Switzerland) (2023)
In September 2017, Hurricane Irma made landfall in South Florida, causing a great deal of damage to mangrove forests along the southwest coast. A combination of hurricane strength winds and high storm surge across the area resulted in canopy defoliation, broken branches, and downed trees. Evaluating changes in mangrove forest structure is significant, as a loss or change in mangrove forest structure can lead to loss in the ecosystems services that they provide. In this study, we used lidar remote sensing technology and field data to assess damage to the South Florida mangrove forests from Hurricane Irma. Lidar data provided an opportunity to investigate changes in mangrove forests using 3D high-resolution data to assess hurricane-induced changes at different tree structure levels. Using lidar data in conjunction with field observations, we were able to model aboveground necromass (AGN; standing dead trees) on a regional scale across the Shark River and Harney River within Everglades National Park. AGN estimates were higher in the mouth and downstream section of Shark River and higher in the downstream section of the Harney River, with higher impact observed in Shark River. Mean AGN estimates were 46 Mg/ha in Shark River and 38 Mg/ha in Harney River and an average loss of 29% in biomass, showing a significant damage when compared to other areas impacted by Hurricane Irma and previous disturbances in our study region.
Keyphrases
  • climate change
  • water quality
  • oxidative stress
  • electronic health record
  • high resolution
  • big data
  • mental health
  • mass spectrometry
  • machine learning
  • diabetic rats
  • particulate matter
  • health insurance
  • deep learning