Toward regional hazard risk assessment: a method to geospatially inventory critical coastal infrastructure applied to the Caribbean

Abstract

Hurricanes and sea level rise pose significant threats to infrastructure and critical services (e.g., air and sea travel, water treatment), and can hinder sustainable development of major economic sectors (e.g., tourism, agriculture, and international commerce). Planning for a disaster-resilient future requires high-resolution, standardized data. However, few standardized approaches exist for identifying, inventorying, and quantifying infrastructure lands at risk from natural hazards. This research presents a cost effective, standardized and replicable method to geospatially inventory critical coastal infrastructure land use and components, for use in risk assessments or other regional analyses. While traditional approaches to geospatial inventorying rely on remote sensing or techniques, such as object-based image analysis (OBIA) to estimate land use, the current approach utilizes widely available satellite imagery and a “standard operating procedure” that guides individual mappers through the process, ensuring replicability and confidence. As a pilot study to develop an approach that can be replicated for other regions, this manuscript focuses on the Caribbean. Small islands rely heavily on a small number of critical coastal infrastructure (airports, seaports, power plants, water and wastewater treatment facilities) and climate related hazards threaten sustainable development and economic growth. The Caribbean is a large and diverse area, and gaps exist between countries in the resources required for planning but much of the region lacks a comprehensive inventory of the land, infrastructure, and assets at risk. Identifying and prioritizing infrastructure at risk is the first step towards preserving the region’s economy and planning for a disaster resilient future. This manuscript uses high resolution satellite imagery to identify and geo-spatially classify critical infrastructure land area and assets, such as structures, equipment, and impervious surfaces. We identified 386 critical coastal infrastructure facilities across 28 Caribbean nations/territories, with over 19,000 ha of coastal land dedicated to critical infrastructure. The approach establishes a new standard for the creation of geospatial data to assess land use change, risk, and other research questions suitable for the regional scale, but with sufficient resolution such that individual facilities can utilize the data for local-scale analysis.

A system for the management of sandy shorelines under climate change: United States Virgin Islands (USVI)

“…A shoreline (line) file (.kml) was created using heads-up digitizing (Becker et al 2021 ) in Google Earth Pro. The geomorphology of the study area was then roughly classified (for 100 m segments) into five categories: sand beach, cobble beach, low cliff/rip-rap walls, medium cliff/small seawalls, and rocky high cliff/seawalls based on the information presented in Hammar-Klose and Thieler ( 2001 ).…”

“…It is worth noting that coastal communities might have digitized building information for zoning and tax purposes. In cases where such data are unavailable, a 'heads-up' digitizing process can be employed using Google Earth Pro, as outlined by Becker et al ( 2021 ) and Bove et al ( 2020 ). Coastal business and residential buildings were used to represent the count of buildings most likely associated with beach recreational activities.…”

Rapid seaward expansion of seaport footprints worldwide

“…While regional and global analyses of risk to seaport infrastructure and international trade networks are becoming more powerful, nuanced, and detailed [2][3][4]13 , current assessments treat the spatial footprints of seaports as static quantities, and do not account for seaport expansion, typically seaward, over time (Fig. 1) 14 .…”

“…Determining from remotely sensed data landward expansion and patterns of change in the total footprints of seaport area across coastal and terrestrial spaces requires a different analytical approach. Nor does our method differentiate among specific uses of seaport-related space, such as terminal facilities, storage, industry, or other integrated layouts 4,13,14 : the seaward footprints that we measure must be interpreted as a partial gauge of gross port area. Imagery from Google Earth and Planet Basemap, and targeted queries in OpenStreetMap, corroborate that the seaward growth we measure at these 65 sites is associated with expansion of seaport complexes.…”