Geology topics

Katherine A. Castagno

Publications and source records attributed to Katherine A. Castagno.

2 recordsLinked to original sources

USGS2022

How much marsh restoration is enough to deliver wave attenuation coastal protection benefits?

As coastal communities grow more vulnerable to sea-level rise and increased storminess, communities have turned to nature-based solutions to bolster coastal resilience and protection. Marshes have significant wave attenuation properties and can play an important role in coastal protection for many communities. Many restoration projects seek to maximize this ecosystem service but how much marsh restoration is enough to deliver measurable coastal protection benefits is still unknown. This question is critical to guiding assessments of cost effectiveness and for funding, implementation, and optimizing of marsh restoration for risk reduction projects. This study uses SWAN model simulations to determine empirical relationships between wave attenuation and marsh vegetation. The model runs consider several different common marsh morphologies (including systems with channels, ponds, and fringing mudflats), vegetation placement, and simulated storm intensity. Up to a 95% reduction in wave energy is seen at as low as 50% vegetation cover. Although these empirical relationships between vegetative cover and wave attenuation provide essential insight for marsh restoration, it is also important to factor in lifespan estimates of restored marshes when making overall restoration decisions. The results of this study are important for coastal practitioners and managers seeking performance goals and metrics for marsh restoration, enhancement, and creation.

Frontiers in Marine Science↗

USGS2021

Labeling poststorm coastal imagery for machine learning: Measurement of interrater agreement

Classifying images using supervised machine learning (ML) relies on labeled training data—classes or text descriptions, for example, associated with each image. Data-driven models are only as good as the data used for training, and this points to the importance of high-quality labeled data for developing a ML model that has predictive skill. Labeling data is typically a time-consuming, manual process. Here, we investigate the process of labeling data, with a specific focus on coastal aerial imagery captured in the wake of hurricanes that affected the Atlantic and Gulf Coasts of the United States. The imagery data set is a rich observational record of storm impacts and coastal change, but the imagery requires labeling to render that information accessible. We created an online interface that served labelers a stream of images and a fixed set of questions. A total of 1,600 images were labeled by at least two or as many as seven coastal scientists. We used the resulting data set to investigate interrater agreement: the extent to which labelers labeled each image similarly. Interrater agreement scores, assessed with percent agreement and Krippendorff's alpha, are higher when the questions posed to labelers are relatively simple, when the labelers are provided with a user manual, and when images are smaller. Experiments in interrater agreement point toward the benefit of multiple labelers for understanding the uncertainty in labeling data for machine learning research.

Earth and Space Science↗