2 Science
TipInteractive Walkthrough
New to the MST scoring pipeline? The Scoring Walkthrough provides a step-by-step interactive visualization of how input rasters are merged, rescaled by ecoregion, and averaged into final sensitivity scores.
2.1 Conceptual Framework
The term (\(V\)) is a function of (\(E\)), (\(S\)) and (\(A\)) (Equation 2.1).
\[ V = f(E, S, A) \tag{2.1}\]
The more exposed and sensitive an area is—and the less able it is to recover—the more vulnerable it is to impacts from offshore human activities. The MST currently focuses on the sensitivity component, quantifying the intrinsic biological sensitivity of marine ecosystems based on the species present and their conservation status. Exposure and adaptive capacity components are planned for future phases (see Chapter 9).
2.2 Cell Vulnerability
For a given cell \(c\), the vulnerability score for a species group \(g\) is the sum across all \(S_g\) species of the product of species presence probability \(p_{sc}\) and extinction risk weight \(w_s\) (Equation 2.2):
\[ v_c = \sum_{s=1}^{S_g} p_{sc} \cdot w_s \tag{2.2}\]
where:
- \(v_c\) = vulnerability score of cell \(c\) for species group \(g\)
- \(p_{sc}\) = probability of species \(s\) presence in cell \(c\) (0–1 scale, from merged species distribution models)
- \(w_s\) = extinction risk weight for species \(s\) (0–1 scale, derived from ESA/MMPA/MBTA/IUCN status; see Chapter 5)
- \(S_g\) = total number of species in taxonomic group \(g\) (e.g., bird, fish, mammal, coral, invertebrate, reptile/turtle)
In other words, if a cell has many species that are both likely to be present and at high risk of extinction, it gets a higher sensitivity score. This helps identify places where rare or threatened species are concentrated.
2.3 Sensitivity Components
The MST assesses sensitivity through three components:
- Species — by taxonomic category (bird, coral, fish, invertebrate, mammal, reptile/turtle): extinction risk weighted species distribution models across 9,819 valid species (see Chapter 3, Chapter 5)
- Habitats — benthic habitats such as coral reefs, seamounts, and hydrothermal vents (planned for future phases)
- Primary Productivity — satellite-derived net primary productivity from the Vertically Generalized Production Model (VGPM), measured as metric tons C km-2 yr-1 (see Chapter 3)
2.4 Ecoregional Rescaling
Raw cell scores vary naturally across regions due to differences in species richness and oceanographic conditions. To enable meaningful comparison, scores are rescaled to a [0–100] range within each BOEM Ecoregion using min-max normalization (Equation 2.3):
\[ v'_c = \frac{v_c - v_{min}}{v_{max} - v_{min}} \times 100 \tag{2.3}\]
where \(v_{min}\) and \(v_{max}\) are the minimum and maximum cell scores for a given component within the Ecoregion.
Scores are relative within Ecoregions, not absolute. A score of 100 represents the most sensitive 0.05° cell for a given component (e.g., fish, marine mammals, or primary productivity) within a given BOEM Ecoregion—not globally. The final score is an equally weighted average of these ecoregionally rescaled components. This means that a score of 30 in Alaska reflects a different ecological context than a score of 30 in the Pacific.
Why rescale by Ecoregion? Without this step, areas with naturally higher numbers of species (like upwelling zones in the Pacific) would almost always get the highest scores, simply because they have more species. While more species may indicate greater ecological importance, it could also be argued that the fewer species in an ecosystem, the more important each is to its function and resilience. Rescaling by Ecoregion prevents scores from being overwhelmed by species-rich areas. However, comparing across Ecoregions would require developing a mathematical relationship between these competing arguments for sensitivity—one that is still a source of earnest academic debate. For this reason, scores are directly comparable among Program Areas within the same Ecoregion, but should not be compared across Ecoregions.
2.4.1 Geographic Scope
The current analysis covers 20 BOEM Program Areas from the 11th National Draft Proposed Program (2025) spanning Alaska, the Pacific, and the Gulf of America (Figure 2.1). The Atlantic OCS region is not included in this program cycle, so ecoregions without Program Areas (e.g., Northeast Continental Shelf, Southeast Continental Shelf, Washington/Oregon) are excluded from the analysis. Species whose distributions do not overlap with any Program Area will not appear in the interactive mapping applications.
The Gulf of America Program Areas—GOA Program Area A (GAA) and GOA Program Area B (GAB)—were derived from the larger Gulf of America Planning Areas and span parts of two Ecoregions: Western and Central Gulf of America (WCGOA) and Eastern Gulf of America (EGOA). Because cell scores are rescaled within each Ecoregion before aggregation, cells in GAA that fall within the WCGOA Ecoregion are rescaled relative to WCGOA min/max values, while cells within EGOA use EGOA min/max values. The aggregated Program Area score is then the area-weighted average across all constituent cells, regardless of which Ecoregion they belong to. This approach ensures that the rescaling reflects local ecological context even when a Program Area spans Ecoregion boundaries. See Chapter 7 for implementation details.
See subsequent chapters for detailed descriptions of: data sources (Chapter 3), taxonomic integration (Chapter 4), extinction risk scoring (Chapter 5), model merging (Chapter 6), and scoring methodology (Chapter 7).