Analyzing Remote Sensing-Derived Normal Difference Vegetation Index to Predict Coastal Protection by Spartina alterniflora

Abstract

Coastal vegetation can provide protection to the coastline through its root structures, which reduce soil erosion, and its stem structures, which dissipate wave energy. The drag a plant induces could be used to quantify the amount of coastal protection that is provided. This study combined field measurements and drone surveys to develop methods for quantifying vegetation drag, focusing on Spartina alterniflora (S. alterniflora), a smooth cordgrass native to the study site: Waquoit Bay National Estuarine Research Reserve. The drag of a single plant is proportional to frontal area. The drag per bed area is proportional to the drag of a single plant and the number of stems per bed area. This study collected plant samples over the growing season to generate allometric relationships between tiller height and individual plant biomass and frontal area, which provides a way to translate remotely-sensed measures of biomass into stem count and frontal area per bed area. The frontal area was measured through digital imaging of individual plants. The elastic modulus of the stem was also measured using an Instron testing machine. For sixteen 1m x 1m test plots, Normalized Difference Vegetation Index (NDVI) extracted from drone multispectral imagery was compared to measured stem count and estimated biomass. The study compared two different years and three time points within a growing season [August 2022; June, August, October 2023]. In addition, at three plots the stem count was manually altered by cutting out 50% and 100% of the plants. This study found that while NDVI can be used to determine the abundance of S. alterniflora, there are several limitations that cause the correlations to be case-specific. Limitations to NDVI-S. alterniflora correlations included: (1) saturation, (2) species inhomogeneity of the area tested, (3) shoot density inhomogeneity of the area tested, and (4) environmental conditions.