Urban Sea Systems
Watershed, urban region, and adjacent sea form one system. Ports, cables, shipping lanes, and data centers make these regions the nodes of global trade and communication, and hot spots of pollution and risk.
Workshop · Seattle · December 2026
The first day of this workshop welcomes all that may be interested in the general ideas involved. The second and third days will be a working gathering that brings together the people who live within the Salish Urban Sea System, those who model it, observe and measure it, and those who may govern it, to explore the challenges and benefits of building an early prototype of Complex Adaptive Digital Twins for the Salish Urban Sea System. There are at least 500 such systems the world over that are integral hubs of the global network of supply chain markets.
Starts in 0 days 0 h 0 min
Why host such a workshop?
They are abundant, and they are hot spots of pollution in both the atmosphere and the ocean. Urban Sea Systems are human-natural ecological enclaves that extend from mountain headwaters through rivers, estuaries, and port cities into the coastal ocean. More than five hundred of them line the world's coasts, and our global inventory lists 583. They enclose a large share of the human population; they organize much of the global economy; and they are governed in sectors that do not match the scales on which water, sediment, organisms, goods, and information actually flow.
Watershed, urban region, and adjacent sea form one system. Ports, cables, shipping lanes, and data centers make these regions the nodes of global trade and communication, and hot spots of pollution and risk.
Currents, rivers, and storms ignore borders. Land use, fisheries, water quality, transport, and emergency management are each managed well in isolation, and rarely together. Cumulative and cascading risks can fall through the gaps.
Shared, continuously updated representations of all interacting sub-systems and conditions let existing institutions easily visualize all important linkages they could not readily resolve before, pose "what if" questions, and cooperate interactively without waiting for governance to be redesigned.
A digital twin, in the sense used here, is not one model or one dashboard. It is a persistent, continuously updated set of domain models, each owned by the experts who already run it, coupled to live observations and to each other, with interfaces that decision-makers can use. The Salish Sea has five decades of observation, modeling, and institutional investment behind it. The workshop asks whether that foundation can be federated into a working prototype, and what a first build would require.
What complex adaptive means
Three ideas, taken in sequence, define the instrument the workshop sets out to build. Each one imposes a design obligation.
Mechanism · John Holland
When many agents interact under local rules, system-level behavior emerges from the couplings and cannot be deduced from the parts.
Couple many domain modules rather than perfect any one. The couplings are the object.
Limit · Simon Levin
Ecosystems are the foremost complex adaptive systems, and their emergent behavior includes outcomes that no model will predict.
Characterize the range of plausible outcomes and let decision-makers rehearse against them. Explore, do not forecast.
Remedy · Elinor Ostrom
Communities can govern shared resources durably through their own evolving rules. The decision-makers belong inside the system.
The parties who must act on the system take part in the instrument that represents it. A module, not a user.
The prototype
The Salish Sea is a large, complex mountain-to-seafloor basin shared by the United States and Canada: the Strait of Juan de Fuca, the Strait of Georgia, and Puget Sound, fed by the Fraser and nine other major rivers, with Vancouver, Victoria, Seattle, Tacoma, and Olympia on its shores. Few coastal regions are as well observed or as densely modeled. Few are governed by as many jurisdictions. That combination makes it the right place to build a first Urban Sea digital twin, and to learn how the pattern could travel to systems at other stages of development.
Think big, start small
A digital twin of an Urban Sea System is not one model. It is a federation of domain modules, each owned by the experts who already run it, coupled through well-defined exchanges of state, flux, and boundary conditions. The workshop starts with seven modules where regional models and committed people already exist, and treats the couplings among them as the real work.
with biogeochemistry
Currents, temperature, salinity, oxygen, carbonate chemistry, and plankton across the estuary, from the Strait of Juan de Fuca to Puget Sound and the Strait of Georgia.
Three complementary operational models, one of which already supplies daily boundary conditions to another across the border.
forcing from the atmosphere
Regional weather ensembles, atmospheric rivers, and downscaled climate projections that drive flooding, runoff, and stormwater design.
land to sea
Forest hydrology, land cover, streamflow, and stormwater from the Fraser, Skagit, Snohomish, and Puget lowland basins into the sea.
The pieces exist across agencies. The workshop must choose the first watershed representation the twin adopts.
nutrients, toxics, microplastics, airsheds
Fate and transport of nutrients, toxics, oil, and microplastics in water, and the thinner strand of airshed modeling for the region.
Water pollution is well covered through module 1. Air pollution needs a named owner.
the human engine
Vessel traffic, rail and truck flows, terminal operations, and their resilience to disruption, on both sides of the border.
Vancouver is already building a port digital twin. The US side needs integration work.
earthquakes, tsunamis, floods, spills
Cascadia megathrust and crustal-fault scenarios, tsunami inundation, coastal flooding, and the human-generated hazards of spills and infrastructure failure.
The human-generated half has no single owner yet. The workshop will decide how much of it the first twin attempts.
the living sea
Salmon, orcas, shellfish, eelgrass and other habitat, and the food webs that link them, across the whole basin and across the border.
Four layers
The seven modules are peers. Beneath them sit two foundations. Above them sits the coupling fabric that turns a set of models into one adaptive system.
Cross-domain flows of water, heat, sediment, nutrients, organisms, decisions, and schedules that bind modules into one system. The hardest and most novel work.
Seven expert-owned models that turn live flow into understanding, each with its own maturity and its own community.
Marine sensing (cabled observatories, moorings, radar, ferry boxes, ships) and civic telemetry (water, sewage, power, traffic, air). A model with no live flow is a study. A model coupled to observation is a twin.
Maps, bathymetry, shorelines, fixed geography, and the historical record. The quiet substrate the system stands on.
Program
Day 1 is open to all and sets out why Urban Seas matter, from hazards to the living sea. Day 2 builds a first specification of how the seven modules connect, and tests it against three worked cascades. Day 3 turns the specification into decisions: who holds the twin, what to build first, and who will carry it.
Draft program. Times, titles, and speakers may change.
Open to anyone who registers. Hazards come first, right after the opening keynote. The morning goes from why Urban Seas matter to the physical system of the Salish Sea. The afternoon covers ports, measurement, and the living sea, then turns to complex adaptive twins and gives the open audience a voice. Day 1 plenaries are recorded, with consent requested.
Speaker to be announced.
Including a report from the DITTO Summit in Yokohama. Speaker to be announced.
Kathryn Sobocinski
Weather in the Salish Sea and the Pacific Northwest (Cecilia Bitz). Watershed dynamics (Brad Butman). Circulation of the Salish Sea (speaker to be announced). The Fraser and the other rivers (Gwyn Lintern).
Paul Blomerus, Clear Seas
Joshua Berger, Maritime Blue
The Internet of Things above sea level (speaker to be announced). Marine measurements (Jan Newton, UW Applied Physics Laboratory, and Jordan Watson, Ocean Networks Canada).
Speaker to be announced.
Followed by questions.
Shiori Sasaki, Esri
Speakers to be announced.
First the room reads one figure together: which of these signs is your work? Then small tables: what should this place look like in 2060, and what would we need to know to get there?
For invited participants. Day 2 builds the specification from two directions. The system lens asks how the region's models must connect. The user lens asks who will use the twin and what they need to know. The two meet in the afternoon's worked cascades. Each session runs in three spaces at once.
The physical system: hydrodynamics and circulation, weather and climate, watersheds and runoff. The living and polluted system: ecological systems, pollution of air and water. The built and hazardous system: ports and supply chains, hazards. Each group corrects and extends a coupling table already filled in from the module summaries.
Real-time sensing above sea level. Sub-sea measurement, from streams and estuaries out to the Pacific. Educational gaming and public use. Each group also names the people who would use its part of the twin.
A Rainier debris flow into the Nisqually valley, as a county emergency manager lives it. Watershed nutrients and a summer heat wave in southern Puget Sound, as a shellfish grower and a public health officer see it. Rising tanker traffic through Haro Strait, underwater noise, and the orcas, as a port operations manager and a whale researcher weigh it.
For invited participants. Day 3 turns the specification into decisions: who holds the twin, what to build first, and who will carry it. It ends with a joint statement. We close at 3:00 so people can travel.
The seven modules' coupling tables, combined overnight into one map of the region's connections.
Natalie Baloy
Who hosts and who governs. Data: sharing, protection, and open formats. What persists: people, funding, and hosting after December.
The room places candidate first projects on a two-by-two of impact and effort.
Shiori Sasaki
Choose one pilot for 2027, sketch the phases beyond it, and name a working group and its members.
The conveners read a draft prepared from Days 1 and 2. The room amends it, and those willing sign.
Participate
Day 1, Tuesday 15 December, is open to anyone who registers. Days 2 and 3 are a working meeting of about fifty invited participants: the modelers, observers, port and utility operators, people from tribes and First Nations, agency staff, and technologists who already work on some part of the Salish system.
Circulation, weather and climate, watersheds, water and air quality, hazards.
Cabled observatories, moorings, radar, ferries, civic telemetry, regional data portals.
Ports, utilities, emergency managers, state, provincial, federal, tribal, and First Nation agencies.
Geospatial platforms, cloud infrastructure, data interoperability, AI and assimilation.
Day 1 (15 December) is open to anyone who registers. Days 2 and 3 are a working meeting of about fifty invited participants. Outputs will be published on this site.
No. The workshop is in person only. Day 1 plenaries are recorded, with consent requested.
Register for Day 1 on the registration page. Days 2 and 3 are by invitation.
Details on costs and travel support will be shared with registered participants.
Yes. A roadmap and a concept paper are planned outputs. Both will be posted here.
Logistics
Room details will be sent to registered participants.
3501 NE 41st Street, Seattle, WA 98105, USA
Campus mapSeattle–Tacoma International Airport (SEA). The Link 1 Line light rail runs from the airport to University of Washington station in about 45 minutes. The Center for Urban Horticulture is about 1.7 km northeast of the station, a 20-minute walk or a short bus or taxi ride.
Amtrak Cascades trains and intercity buses connect Vancouver, BC to Seattle in three to four hours, arriving at King Street Station. A passenger ferry links Victoria's Inner Harbour to Pier 69 on the Seattle waterfront. From downtown, take the Link 1 Line north from International District/Chinatown or Westlake station to University of Washington station.
Hotel options include rooms at the Silver Cloud Hotel, 5036 25th Avenue NE, Seattle, about 1.6 km from the Center for Urban Horticulture. More options will be shared with registered participants.
Conveners
The workshop is convened by ocean scientists and the organizations that observe the Salish Sea, with an organizing committee drawn from the University of Washington and partners on both sides of the border.
Convener
Professor Emeritus, School of Oceanography
University of Washington
Oceanographer and lead architect of the Regional Cabled Array, the interactive ocean observatory off the Pacific Northwest coast. Co-author of the Urban Sea Systems concept.
Convener
Ocean Networks Canada / University of Victoria
Ocean Networks Canada operates the cabled ocean observatories off British Columbia, including arrays in the Strait of Georgia and Saanich Inlet.
Convener
Marine geologist, Center for Habitat Studies
Moss Landing Marine Laboratories, San José State University
Marine geologist and seafloor habitat mapper. Co-author, with John Delaney, of the 2025 characterization of Urban Sea Systems.
Opening keynote
Former Administrator
National Oceanic and Atmospheric Administration
Opening keynote on Day 1: Urban Seas, digital twins, and the blue (and new blue) economy.
Organizing committee
Associate Professor, Earth and Space Sciences
University of Washington
Seismologist working on data-driven monitoring of geohazards and on digital twins of weather-compounded hazards. Leads the GAIA HazLab.
Organizing committee
Professor, School of Oceanography
University of Washington
Marine seismologist. Studies ocean volcanoes, uses seafloor seismic networks to follow baleen whales, and works to improve seismic and geodetic monitoring of the Cascadia subduction zone.
Organizing committee
Postdoctoral Scholar, COSZO
University of Washington
Seismologist working on Cascadia Offshore Subduction Zone Observatory data flow, data QA/QC systems, and agentic AI workflows for geophysics.
Support
The workshop is convened by the University of Washington and Ocean Networks Canada, with advice and support from partners on both sides of the border.
Partner list in progress. Organizations appear here once their support is confirmed.
Resources
Start here to arrive with a shared vocabulary.
Defines Urban Sea Systems and characterizes three North American examples: the Salish Sea, Chesapeake Bay, and the south shore of Oahu.
The argument behind the workshop. Copies will be shared with participants.
Names a Healthy Urban Sea as one of the cross-cutting themes for U.S. contributions to the UN Decade of Ocean Science for Sustainable Development.
The international effort to build digital twins of the open ocean, the complement to the coastal and urban problem addressed here.
An Urban Sea System today. Watersheds, cities, ports, and coastal waters form a single coherent, dynamically coupled system that is economically powerful, ecologically challenged, and episodically threatened by natural or human-induced hazards.
J.R. Delaney, H.G. Greene and M.R. Delaney
Urban Sea Systems line the world's coasts and form the nodes of a global trade network. This map marks more than two hundred of them.
J.R. Delaney and H.G. Greene, after Greene and Delaney (2025)
The dual-layer structure of global trade. Shipping routes at the surface and submarine cables on the seafloor converge at the ports of Urban Sea Systems.
J.R. Delaney, H.G. Greene and M.R. Delaney
The Salish Sea and its watershed, shared by the United States and Canada. The upper Fraser basin more than doubles the conventionally drawn watershed.
J.R. Delaney, H.G. Greene and M.R. Delaney
The six modules first proposed for a complex adaptive digital twin of an Urban Sea System. Ecological systems has since been added as a seventh. Each module is a domain model owned by experts and coupled to the others.
J.R. Delaney
Layered architecture for the Salish Urban Sea digital twin. Baseline and observing layer beneath, domain modules above, and the seams that couple them.
J.R. Delaney
A digital twin as enabling infrastructure. Interconnected modules, one dynamic system, and outcomes for the people who use it.
J.R. Delaney
Three North American Urban Sea Systems: the Salish Sea, San Francisco Bay, and Chesapeake Bay.
J.R. Delaney and H.G. Greene
The same coast as a natural refuge, before it became a port-centered region. What will it look like in thirty years?
J.R. Delaney, H.G. Greene and M.R. Delaney
A continuous, adaptive cycle. Observations feed the twin, models and simulations generate insights, and insights inform the decisions and interventions that reshape the real world.
J.R. Delaney, H.G. Greene and M.R. Delaney
From theory to design obligations. Holland's mechanism, Levin's limit, and Ostrom's remedy each require a feature of the instrument: federated, bounded rather than predictive, and governed from within.
J.R. Delaney, H.G. Greene and M.R. Delaney
Real world, twin, and governance. The real world informs the twin, governance reshapes what is modeled and how, and interventions act on the system.
J.R. Delaney, H.G. Greene and M.R. Delaney