Increases in sea-level rise (SLR), nuisance flooding, and changing storm patterns in coastal areas are raising awareness of the need to mitigate, plan, and consider alternatives in construction guidelines for the safety of future and planned construction and human health and safety [1
]. Varied approaches have been developed to identify and map such coastal hazards for coastal planners and decision-makers [5
]. However, few studies account for the combined effects of SLR and storm-driven coastal flooding on the local scale across vast geographic expanses; even fewer studies account for non-stationary changes in projected water levels and their resulting exposure hazards and socio-economic impacts [7
]. To address this void, the Coastal Storm Modeling System (CoSMoS) was developed to provide planners, managers, policy-makers, and engineers with local-scale (approximately 10–100 m) data on probable future coastal exposure hazards across large geographic scales (approximately one hundred to several thousand kilometers) [13
The third-generation CoSMoS model and its application in Southern California (USA), using a mid-emissions climate scenario (representative concentration pathway (RCP) 4.5), are presented in Part 1 [16
] of this two-part manuscript. More extreme wave climate conditions are illustrated for California in the RCP 4.5 scenario [17
] and, accordingly, it is used for detailed hazard simulations in CoSMoS [16
]. The CoSMoS framework projects global changes, which are driven by Global Climate Models (GCMs) to local scales via a suite of regional and local scale models simulating coastal hazards in response to projections of 21st century waves, storm surges, anomalous variations in water levels, river discharge, tides, and SLR. A detailed discussion of the methodology, modeling framework, recent improvements, model validations/limitations, and an incorporation of uncertainty into coastal hazard projections can be found in Part 1 [16
In this second part of the manuscript, results of the modeled hazards are presented and conjoined with land cover, population statistics, and socio-economic data to provide 21st century hazard-exposure estimates along the largely developed Southern California coastline, a region that thrives on tourism, software, automotive, ports, finance, and biomedical industries, contributing to more than 50% of California’s Gross Domestic Product (GDP), ranked the fifth largest worldwide [18
The hazards of interests in this study are coastal erosion and flooding. Exposure refers to the presence of various societal elements (e.g., people, buildings, resources, critical facilities, and infrastructure) that are in hazard zones, and therefore susceptible to damage or loss. Vulnerability describes the characteristics of individuals and assets as well as larger socioeconomic factors that influence the degree to which an individual, system, or community is susceptible to the damaging effects of a hazard. For example, although a large area of residential housing may be equally exposed to coastal flooding, the vulnerability of individual households will vary due to demographic characteristics of homeowners (which influence one’s ability to prepare for and mitigate potential losses) and due to differences in the types of structures (which influence the ability to withstand impacts). In addition, two adjacent towns may have equal hazard exposure, but their overall vulnerability to flooding varies if the number of homes exposed to flooding represents 5% of available housing in one town and 95% of housing in the other town.
To communicate coastal hazards, exposures, and vulnerabilities as well as assess the socio-economic impacts, the data are made available via two publicly accessible web tools (Figure 1
). “Our Coast, Our Future” [19
] (OCOF; www.ourcoastourfuture.org
) is a web application for data visualization, synthesis, and access to all output products from the CoSMoS model. The OCOF mapping interface provides coastal managers and the general public with a user-friendly means to visualize how various SLR scenarios alone and in combination with three different future return-period coastal storms are projected to flood or erode. Users can export summary tables and reports detailing changes in flood extent by scenario on a scale relevant to local planners. The Hazard Exposure Reporting and Analytics [20
] (HERA; https://www.usgs.gov/apps/hera/
) web tool translates the flooding hazards into community-based exposure statistics and quantifies populations, property, and critical infrastructure at risk in terms of exposure statistics and monetary values on a community level.
The scientific methods that underpin the projected hazards are based on state-of-the-art science that includes many of the latest developments and understandings of coastal processes (i.e., non-linear effects of currents and waves, reflection, refraction, and blocking of wave energy due to complex bathymetry; see Barnard et al. and O’Neill et al. [15
]) making it difficult to communicate assumptions, limitations, as well as the strength and value of the data to non-technical and non-specific science-educated audiences. To address these concerns, both traditional and innovative stakeholder engagement, training, and outreach efforts have been tested. Although the translation of the science remains challenging, we have pinpointed several tactics that are likely to be useful in similar large-region, high-resolution studies elsewhere.
The aim of this paper is to (1) highlight the need to account for dynamic water levels in addition to static SLR for estimating future coastal flood hazards and vulnerabilities along high-energy coastal environments such as Southern California, USA and (2) to present means and introduce innovative approaches for conveying the hazards, socio-economic impacts, and underlying scientific basis to a broad audience including coastal managers, planners, and engineers. The remainder of this manuscript describes the data and methods used to quantify exposures and vulnerabilities along the Southern California coast. Results pertaining to erosion and flood hazards are presented with a particular emphasis on the added risk when storms are accounted for in addition to SLR. Results of the socio-economic impact analyses are then presented for a select set of assets and demographics rather than all available results to illustrate the use and applicability of stakeholder user-tools. The final sections present strategies and innovative outreach activities for dissemination of the information as well as a summary and discussion of findings.
The methods developed for simulating hazards with the latest generation of CoSMoS are outlined in Part 1 of this manuscript [16
], which include presentation and discussion on models, selection of storm conditions, model validation/limitations, and uncertainty. The latest iteration of CoSMoS is implemented in Southern California, an active and complex tectonic region spanning over 500 km and five counties (Figure 1
). The coastal landscape is generally characterized as beaches backed by semi-resistant bedrock sea-cliffs as well as coastally constrained estuaries and low-lying areas at the foot of coastal mountain ranges (see Part 1 for more details). CoSMoS-modeled hazards for Southern California include outputs of coastal erosion, wave heights, wave runup, total water levels, current speeds, flood extents, and flood depths/durations for 40 ‘scenarios’ consisting of all combinations of 10 SLR elevations (0–200 cm SLR in 25 cm increments, plus 500 cm), three coastal storm intensities (annual, 20-year, and 100-year), and a no-storm condition [16
]. Coastal erosion outputs included management scenarios involving beach nourishment and the existence and maintenance of hard structures (see Section 3.1.1
). Low-lying flood-prone areas and uncertainties in flood extents and shoreline change are also generated as part of the model output.
Resulting model data were converted to static GeoTIFF rasters (flood depth, wave height, and current velocity), polygon shapefiles (flood extent, low-lying areas, and uncertainty) or point shapefiles (wave runup and shoreline change) and were processed for the OCOF cyberinfrastructure to display and provide exposure hazard map data. The cyberinfrastructure was built on the Open Source Geospatial Foundation stack of software. Simple raster tiles were first rendered from the GeoTIFF data layers; point and polygon layers were loaded into PostgreSQL (an open-source database; version 9+; available https://www.postgresql.org/
)/PostGIS (an open-source, GIS-support software program; version 2+; available https://postgis.net/
) database and piped through the GeoServer web service for rendering and display on the map (see Appendix A
for a list of terms, acronyms, and software platforms). Initially, the data were provided for download as large zip files but these proved to be too cumbersome for many users; a re-organization of the data tiled by scenario, output product (e.g., flood depth/extent, duration, wave height), and individual counties has improved users’ ability to access the data and increased overall user satisfaction.
Geospatial data summarizing various population, business, land cover, and infrastructure were used to estimate community exposure to a given flood hazard zone in HERA [20
]. Residential populations were estimated using block-level population counts compiled from the 2010 US Census [23
]. Demographic and economic factors, such as age, health, ethnicity, race, and health, and tenancy, can amplify an individual’s sensitivity to hazards [24
]; therefore, the 2010 block-level data were used to estimate demographic attributes related to these socio-economic indicators of sensitivity.
Business populations and regional trends of exposure were estimated in HERA using employee counts organized by North American Industry Classification System (NAICS) codes [23
] at individual businesses using a georeferenced, proprietary employer database [26
]. Business types based on NAICS codes were generalized in this analysis into five classes: (1) government and critical facilities; (2) manufacturing; (3) services; (4) natural resources; and (5) trade.
Land cover indicators include the amount and type of land in hazard zones based on 30-m-resolution data extracted from the 2011 National Land Cover Database (NLCD) [27
]. The HERA application currently focuses on land classified in NLCD as either wetlands or developed.
Hazard exposure of critical facilities and infrastructure was estimated using the length of rail and road networks (infrastructure) and the number of schools, medical facilities, police stations, and fire stations (facilities). These facilities are considered critical because they provide public safety services or house vulnerable populations. Data sources for critical facilities and infrastructure include a wide array of county and federal sources [28
For each variable, geographic information system (GIS) software was used to overlay data representing community boundaries, the community indicator, and a specified flood hazard zone. Two variables for each asset were estimated at the community level: (1) a total amount (or length, in miles, for road and rail networks) of an asset in a hazard zone and (2) a community percentage. For resident and employee populations in hazard zones, the community percentage reflects the exposed amount compared to the total amount within a community. For the business types, percentages reflect the number of businesses of a certain type divided by the total number of that business type in the community. For the demographic attributes, community percentages reflect the percentage of a specific demographic attribute relative to the total number of residents in the hazard zone, not the community total. Spatial analysis of vector data focused on determining if points (businesses and critical facilities), lines (roads and rails), or polygons (census blocks) were inside hazard zones. If census-block polygons overlapped hazard polygons, final population values were adjusted proportionately using the spatial ratio of each sliver within or outside of a hazard zone.
4. Stakeholder Engagement and Outreach
Outreach and engagement with planners, engineers, emergency managers, and environmental scientists from coastal cities, counties, utilities, state agencies, non-governmental organizations, and the private sector were conducted both in advance of and following the release of model results. Outreach was designed and delivered in collaboration with established and trusted regional partners or networks to ensure local relevance. Workshops were held prior to the development of web tools to gain an understanding of what type of data, formats, and displays might be most suitable for end-users. Once model results were complete and incorporated into the web tools, high-level trainings and demonstrations of the OCOF and HERA web tools were conducted. To bolster interest and use, trainings were tailored to the specific needs and interests of local stakeholders and immediate access to the data in their respective areas was provided. For instance, the San Diego County workshop included a panel highlighting local projects engaged in SLR planning as well as a separate session that focused on model details, assumptions, and limitations for more technical end-users. For the Los Angeles County workshop, community planning exercises were conducted where attendees could view the SLR and erosion projections on paper maps and brainstorm adaptation ideas. Over the course of the Southern California project, we participated in eight workshops over three years, reaching over 500 participants across all Southern California counties.
In addition to the more traditional outreach and engagement strategies, virtual reality (VR) and 360-degree 3-D videos were used for communicating future coastal flood hazards. For instance, in the city of Santa Monica, California, projected flood extents were used to create virtual images of the beach under different states of SLR, both with and without the effects of coastal storms (see mobileowl.co/samo; Figure 8
a–c). Possible adaptation strategies were also presented in some locales. Residents and visitors were guided through a series of images in a virtual-reality viewer that showed the beach
flooded due to a 100-year storm modeled with CoSMoS,
in the future with 2 m of SLR (Figure 8
in the future with 2 m of SLR and the 100-year coastal storm, and potentially
modified with a possible adaptation strategy (Figure 8
As users moved through the images, they were asked a series of questions on demographics and levels of concern as the flood hazards increase. These images were available via an in situ platform, nicknamed the “owl” (by Owlized™, see Figure 8
a inset), that was placed on Santa Monica Pier from November 2016 to January 2017; the “owls” were used to support Santa Monica’s local coastal planning and outreach efforts. The augmented images within the viewer provided visceral opportunities to visualize the complex scientific information used by the Santa Monica community in its planning. The visualizations are still available via the online mobile viewer (mobileowl.co/samo) and continue to be viewed by interested stakeholders.
A second innovative technology and approach using VR allows users to visualize how coastal areas may be effected everyday under future SLR conditions. A video system developed in partnership with Google™ and consisting of 16 GoPro™ cameras (the GoPro Odyssey™) was used to film 360-degree videos that show beaches during the highest (‘King’) tides of the year (Figure 8
d–f). The videos were uploaded to YouTube and, using VR headsets, residents can explore and observe their local beach during these extreme conditions, visualizing how it might appear under normal tide conditions in the coming century if sea level continues to rise. Both sets of VR visualizations (“owl” and King tide videos) have been used for education and outreach purposes to help make complex scientific information more accessible. They can both be accessed via mobile platforms as well as via home computers with sophisticated VR headsets (such as any Google™ VR viewer).
5. Discussion and Summary of Findings
The overarching concept of CoSMoS is to leverage projections of global climate patterns over the 21st century from recent Global Climate Models (GCMs). Coarse resolution GCM projections are downscaled to the local level and used as boundary conditions to sophisticated ocean modeling tools that simulate complex physics to accurately predict local coastal water levels and flooding for the full range of expected SLR (n = 10: 0–2 m in 0.25 m increments and 5 m) and storm scenarios (n = 4: average daily/background conditions, annual, 20-year, and 100-year). Resulting model projections include spatially explicit estimates of flood extent, depth, duration, uncertainty, water elevation, wave run-up, maximum wave height, maximum current velocity, and long-term shoreline change and cliff retreat.
The model system produces coastal hazard projections suitable to aid local climate adaptation planning. The results are provided to the public via two heavily vetted and user-tested web tools, one that presents the hazards in a map-style interface, (“Our Coast, Our Future” (OCOF): www.ourcoastourfuture.org
) and a second one that integrates the hazards with geospatial demographic and socio-economic data to provide information on exposures and vulnerabilities (Hazard Exposure Reporting and Analytics (HERA): https://www.usgs.gov/apps/hera/
). Both tools have the dual benefit of providing user-friendly web tools that allow the interested public to explore complex scientific information, similar to how they would use a web browser to explore a local map, as well as providing robust scientific information that can be used by municipal, county, and statewide planners and managers in their coastal adaptation and local hazard mitigation plans. Additionally, the web applications can be used by diverse audiences for multiple purposes. For those coastal communities that do not have access to geographical information system (GIS) specialists, OCOF and HERA allow these communities to explore the full suite of SLR and storm hazards and impacts to incorporate into their planning efforts. For communities with access to technical GIS capabilities, web applications are used as a public outreach tool that allows interested residents and community members to explore the scientific information to supplement their own understanding. Thus, they play an important education/outreach function supporting local coastal adaptation planning.
Using a continuous time series of nearshore wave conditions as well as storm surge and sea level anomaly levels in combination with SLR, the coastline change models developed for this study indicate a spatial average of ~25–40 m of beach erosion and ~10–85 m cliff retreat in Southern California [30
] (rounded to the nearest 5 m). This amount of shoreline retreat would completely erode as much as 67% of the beaches in Southern California [30
]. Lower SLR scenarios result in less but not insignificant erosion; for example, 50 cm of SLR results in an average of 15 m of beach loss and 10 m of cliff retreat.
Flood hazards modeled with the deterministic, dynamic CoSMoS model and including projected coastline change, show that, across Southern California, 100 to 200 cm of SLR will inundate 40–150 km2
of land. It also shows that the effects of a 100-year coastal storm would flood an additional 54% (150–230 km2
) to 129% (40–340 km2
) of land area. More near-term projections of 25 cm SLR (by approximately the year 2030) [32
] are estimated to permanently flood ~10 km2
, with an intermittent flood extent increasing the area affected by nearly 350% (10–45 km2
) when the 100-year storm is also taken into consideration. The results demonstrate that, if sea level continues to rise, many areas will be impacted by flooding in both the long- and short-term, and that storm conditions, combined with even small amounts of SLR expected within just a few decades, will substantially increase the exposure hazard.
Translated to socio-economic impacts, 25–200 cm of SLR places ~20,000–164,000 residents at risk of being permanently flooded along the Southern California shores. Building replacement values are estimated to be between $3.64 billion and $26.10 billion (2006 value, unadjusted for inflation). Accounting for the 100-year storm exposes an additional 56–109% of residents and increases building replacement costs by 46% to $38.2 billion, thus highlighting the importance of including storms in vulnerability assessments.
For actual implementation of hazard mitigation or climate adaptation actions, quantified projections of impacts at the community level are invaluable. The HERA tool provides analytics summarized in the form of maps and graphs for evaluation of vulnerable areas, populations, infrastructure, and economic sectors as well as the ability to download all the data for off-line in-depth local-scale analysis and planning. Using this tool, social vulnerability at the community level can be evaluated according to relative distribution of income, race, age, access to resources, viability of critical infrastructure (e.g., hospitals, roads), building replacement costs, and diversity of economic assets [24
To communicate the availability, uses, and implications of the modeled hazards, exposures, and vulnerabilities, numerous workshops and outreach activities, tailored to the specific needs and interests of local stakeholders and delivered through existing, trusted networks and partnerships have been and continue to be held in regions where the model system has been applied. At each workshop, an overview of the CoSMoS model and regional results are provided, demonstrations and hands-on trainings of the web tools are conducted, and access to data specific to the region is highlighted and discussed. In addition to the more traditional outreach and engagement strategies, virtual reality technology and activities are being developed and applied to reach greater audiences and to better communicate future coastal flood hazards. CoSMoS, and the associated web tools OCOF and HERA, are used for local and state-level coastal planning as well as hazard mitigation planning by approximately 30 coastal cities and counties in California. It is also utilized by many of the state agencies, such as the California Coastal Commission, California Department of Emergency Services and the California Department of Transportation, nongovernmental organizations, and regional-scale collaborations. Although both the modeling methods and outreach activities continue to evolve, the success of providing useful information for coastal planners, engineers, and other stakeholders is underpinned by close relationships and ties with regional partners and local stakeholders who help envision, build, and develop effective products and tools for the critical end-user.