Semester of Graduation

Spring 2026

Degree

Master of Science in Civil Engineering (MSCE)

Department

Civil and Environmental Engineering

Document Type

Thesis

Abstract

Coastal Louisiana’s shorelines are rapidly changing due to subsidence, sea-level rise, storm impacts, and sediment dynamics. This study develops and evaluates a semi-automated shoreline extraction toolbox (SE+) that applies a Modified Normalized Difference Water Index (MNDWI) segmentation framework for long-term shoreline monitoring along the Louisiana Chenier Plain. Annual Landsat surface reflectance composites (2010-2024) were processed using Gaussian smoothing, adaptive Otsu thresholding, and morphological filtering to determine water-land borders and generate vector shorelines. Each run produced a metadata JSON file to document processing settings and ensure reproducibility and transparency. Shoreline positional uncertainty was quantified using a root-sum-square technique that combined pixel-related uncertainty, tidal/stage variability, and georeferencing error, with the georeferencing term constrained using a 2017 LiDAR-derived DEM. Shoreline change rates were calculated using the Digital Shoreline Analysis System (DSAS) with baseline and transects constructed at 100 m intervals and 1000 m in length. Three DSAS metrics were utilized to calculate change rates: End Point Rate (EPR), Linear Regression Rate (LRR), and Weighted Linear Regression (WLR). The findings indicate that coastline retreat dominates the study area. The mean change rates were -3.75 m/yr (EPR), -4.14 m/yr (LRR), and -4.17 m/yr (WLR), with erosional classifications occurring along 76-78% of transects. Strong agreement amongst metrics suggests that uncertainty-aware regression yields rates comparable to unweighted regression while enhancing statistical robustness through formal uncertainty weighting. Overall, the SE+ methodology offers a repeatable method for regional-scale shoreline mapping and supports reliable quantification of long-term coastal change along the Louisiana Chenier Plain.

Date

4-10-2026

Committee Chair

Abdalla, Ahmed

LSU Acknowledgement

1

LSU Accessibility Acknowledgment

1

Download

Available for download on Monday, March 26, 2029

Share

COinS