Is this project an undergraduate, graduate, or faculty project?
Undergraduate
Project Type
individual
Campus
Daytona Beach
Authors' Class Standing
Morgan Grayston, Senior
Lead Presenter's Name
Morgan Grayston
Lead Presenter's College
DB College of Aviation
Faculty Mentor Name
Daniel Burow
Abstract
When tornadoes hit wooded areas, analyzing the frequency of tree falls is a valuable tool in estimating tornado wind patterns and intensity. Often, tree falls are identified using imagery obtained via unpiloted aerial systems (UAS). However, this requires having the resources to obtain this imagery of the damaged region, which is not always possible. Additionally, digitizing thousands of tree falls is a very time-intensive process. One alternative is to use high resolution (<5 >m) multispectral satellite imagery to estimate where tree falls occurred. This study compares spectral indices from Planet satellite imagery to tree falls manually digitized from imagery obtained via UAS at a location in Jacksonville, Alabama, that was hit by an Enhanced Fujita (EF)-3 tornado on 19 March 2018. Using this satellite imagery, we examine changes in indices such as Normalized Difference Vegetation Index (NDVI) and Enhanced Vegetation Index (EVI) at locations of tree falls digitized on the UAS imagery. The long-term goal is to inform future researchers on how to accurately estimate tree falls without the need to obtain UAS imagery.
Did this research project receive funding support (Spark, SURF, Research Abroad, Student Internal Grants, Collaborative, Climbing, or Ignite Grants) from the Office of Undergraduate Research?
No
Using multispectral imagery to estimate tornado-induced tree fall patterns identified via UAS
When tornadoes hit wooded areas, analyzing the frequency of tree falls is a valuable tool in estimating tornado wind patterns and intensity. Often, tree falls are identified using imagery obtained via unpiloted aerial systems (UAS). However, this requires having the resources to obtain this imagery of the damaged region, which is not always possible. Additionally, digitizing thousands of tree falls is a very time-intensive process. One alternative is to use high resolution (<5>m) multispectral satellite imagery to estimate where tree falls occurred. This study compares spectral indices from Planet satellite imagery to tree falls manually digitized from imagery obtained via UAS at a location in Jacksonville, Alabama, that was hit by an Enhanced Fujita (EF)-3 tornado on 19 March 2018. Using this satellite imagery, we examine changes in indices such as Normalized Difference Vegetation Index (NDVI) and Enhanced Vegetation Index (EVI) at locations of tree falls digitized on the UAS imagery. The long-term goal is to inform future researchers on how to accurately estimate tree falls without the need to obtain UAS imagery.