Utilizing a Geospatial Foundation Model for Coastline Delineation in Small Sandy Islands

TL;DR

This paper evaluates NASA and IBM's Prithvi-EO-2.0 geospatial foundation model for shoreline detection on small islands, demonstrating high accuracy with minimal training data and highlighting its potential for coastal monitoring in data-scarce areas.

Contribution

It provides an initial assessment of Prithvi-EO-2.0's effectiveness in shoreline delineation and introduces a new labeled dataset of satellite images of small islands.

Findings

High performance with as few as 5 training images (F1 0.94, IoU 0.79)

Strong transfer learning capabilities of Prithvi-EO-2.0

Potential for supporting coastal monitoring in data-poor regions

Abstract

We present an initial evaluation of NASA and IBM's Prithvi-EO-2.0 geospatial foundation model on shoreline delineation of small sandy islands using satellite images. We curated and labeled a dataset of 225 multispectral images of two Maldivian islands, which we publicly release, and fine-tuned both the 300M and 600M parameter versions of Prithvi on training subsets ranging from 5 to 181 images. Our experiments show that even with as few as 5 training images, the models achieve high performance (F1 of 0.94, IoU of 0.79). Our results demonstrate the strong transfer learning capability of Prithvi, underscoring the potential of such models to support coastal monitoring in data-poor regions.