The Utilization of Regression and Analysis of Variance to Model Hurricane Storm Surge Sedimentation on Coastal Marshes along the Northern Gulf of Mexico Coastline

Abstract

<p>Coastal marshes along the northern Gulf of Mexico coastline provide very important ecosystem services such as serving as habitat for a variety of flora and fauna and providing flood protection for inland areas. A growing body of research has documented how hurricane storm surge sedimentation has increased the elevation of coastal marshes along the northern Gulf of Mexico coastline. This study investigates spatial variations in sediment distribution on McFaddin National Wildlife Refuge, Texas, USA, which is in the geographic region that was impacted by the right-front quadrant of Hurricane Ike. This research builds upon a prior study on hurricane storm surge sedimentation in which the sediment deposits from hurricanes’ Audrey, Carla, Rita, and Ike were identified on a marsh transect on McFaddin National Wildlife Refuge. The purpose of this study was to discover how hurricane storm surge sedimentation spatially varies in relation to the landfall location of Hurricane Ike. Fieldwork conducted in 2017-2018 involved digging shallow pits on four coastal marsh transects between Sabine Pass, Texas and High Island, Texas. Elevations were measured at each pit site along all four transects using a telescopic lens and stadia rod. The transects extend 880-1630 meters, with pit sites beginning near the coastline and extending landward. Results obtained in the field indicate that the Hurricane Ike sediment deposit has been found on all four transects, and that the deposit decreases in thickness moving landward along each transect. Furthermore, the observational results of this study were used in Regression Analyses to model hurricane storm surge sediment deposit thickness based on pit site distance inland, pit site elevation, and distance from the landfall of Hurricane Ike. Moreover, Analysis of Variance revealed whether distance inland, distance from landfall location, and the interaction between distance inland and distance from landfall location had any significant effect on storm surge deposit thickness. Actual sediment deposit thicknesses measured in the field were compared to the Regression and Analysis of Variance results. Results show that the Power Law Curve from the Regression Analyses was the most robust predictor of pit site sediment thickness based on distance inland, with an R<sup>2</sup> value of 0.538. Additionally, the Regression and Analysis of Variance results revealed that transect distance from the landfall location of Hurricane Ike was the only independent variable that could not predict or explain storm surge deposit thickness; which is very likely due to all four transects being in the right-front quadrant of landfalling Hurricane Ike. The findings of this study provide improved understanding of the spatial relationship between storm surge sedimentation and storm surge heights, valuable knowledge about the sedimentary response of coastal marshes subject to storm surge deposition, and useful guidance to public policy aimed at combating the effects of sea-level rise on coastal marshes along the northern Gulf of Mexico coastline.</p><p> </p>