Genies and bottles: Louisiana's master plan to save its coastline

Louisiana loses a football field worth of coastline every 45 minutes. The state is the largest funding source for coastal restoration and resilience. Read about their $50B plan to reinvest in the coast, with a particular focus on what's possible and the gaps that are left to community funding.


The Louisiana Coastal Master plan is published every six years as an assessment of the present and future state of the Mississippi River Delta, its outlying coastal biomes, and the communities that live there. It also details the state's intended roadmap for coastal protection and restoration. The 2023 plan evaluated 131 projects for inclusion in a roadmap totaling $50B in investment over the coming 50 years. The disappearance of Louisiana's coastal biomes is well documented, the simple version of that story is that the state loses a football field of land every 45 minutes. Rarely, however, are the challenges presented in so condensed a space. The coastal protection efforts, when viewed collectively, represent one of the largest environmental engineering projects in the world.

For its scale, the master plan receives relatively little federal funding. Of the $500M-$1B Louisiana spends on coastal protection every year, a majority comes from Deepwater Horizon settlements and state appropriations. In 2021, the federal government made a major allocation to coastal protection, investing $2.5B in major projects, but the state level has been (and will continue to be) the authoritative account of these efforts.

Points that stuck with me

the coastal master plan is a massive undertaking

The Coastal Protection and Restoration Agency -- the publisher of the report -- was founded in 2005 post Katrina + Rita. Since that time, CPRA and its partners have secured $21.4 billion in pursuit of coastal restoration and risk reduction efforts. Over that time, CPRA has completed more than 140 projects to benefit coastal Louisiana, with:

  • 55,807 acres benefited (87.2 sq mi)
  • 193M cubic yards of sediment placed
  • 369 miles of improved levees
  • 71.6 miles of restored barrier islands

The scope of the problem is massive too with 7,700+ miles of coastline and 2M Louisianans at risk.

the outcome of this plan is highly dependent on total sea level rise

The Coastal Master Plan judges is program on the amount of land it preserves and the total cost of economic damage avoided over the coming 50 years. It's model assume two possible futures from a broad climate perspective, both broadly consistent with the IPCC's report on the climate that would result from limiting atmospheric temperature rise to 1.5-2 degrees Celcius.

  • The low case assumes 1.6 ft of sea level rise is expected over the next 50 years, and increased storm intensity of 5%
  • The high case assumes sea level rise of up to 2.5 ft, higher subsidence, and increased storm intensity of 10% in 50 years.

Assuming relatively conservative climate change and no additional resilience spend, Louisiana is on pace to lose nearly 1,000 square miles of land over the next fifty years.

these are permanent consequences

The Coastal Master plan is exceptionally ambitious (and could always be more so with additional federal dollars). The outcome, however, recognizes that there will be no going back to the way that things were for Louisianans. From the start, the report recognizes, "we cannot recreate the coast of the 20th century or even retain the coast of today resilience, not victory is the goal."

In the low climate case, assuming full program deployment over 50 years, a very different Louisiana would face less hurricane and tropical storm risk than it does today. In the high climate case, even with the action considered, there are going to be a significant number of communities for whom voluntary repurchase is the only state-approved course of action. Climate impacts are here and some of them are going to be permanent.

the language of natural disasters is changing

Previously, 100-year flood language would have described incidents with a 1% probability of occurring in a given year. Coastal Louisiana has experienced back to back 100-year flood events. The report refers instead to Annual Exceedence Probability (AEP), and this is the standard against which much of product efficacy is judged.

the master plan recognizes that it can't fully control implementation at the local level

This is not a new revelation, but it's made explicit in plan language and incentives. To do so is necessary for much of the plan's success. Many federal programs (including flood insurance) have struggled because they relied on municipal zoning policy changes and enforcement that never materialized.

If the report recognizes that local political structures can impede project rollout, however, it does less to recognize where local structures could accelerate rollout. There are significant listening sessions in which community perspective was solicited, but the vast majority of restoration spend (85%) is reserved for centrally-designed or delivered programs.

Additionally, a vast majority of the $11B in non-structural risk reduction funds (floodproofing, voluntary acquisition, building elevation) are allocated through local structures, but they impose an opportunity cost on all other projects that becomes prohibitive. For example, the model forces most communities to select either non-structural or structural risk reduction, and mandates that any structural risk reduction project on the coast outperform all non-structural risk reduction on the coast. While this emphasizes the importance of non-structural risk reduction (and highlights the reluctance of many communities to undertake it), it also takes significant agency away from communities that might benefit from a portfolio of mitigation strategies.

the map is not the terrain

I puked a little writing that, but in this case, it's literally a map and a terrain. Model reflexivity is significant in one of the world's most complex adaptive ecosystems. The Mississippi River Delta will convince you there is a god. The model required to simulate its behavior will convince you there is a hell, too. These models are specific to 30m x 30m squares. That "m" stands for meters, not miles. They are run on supercomputers. They're driven by larger than 30m x 30m inputs, put the precision of the model is a feat of human engineering. Readers of the report should consider the projects selected as dependent on four interrelated simulations:

  1. The landscape model of Louisiana (Integrated Compartment Model or ICM) predicts coastal hydrology, wetland morphology, vegetation dynamics, and the suitability of habitats to support an array of fish and wildlife. This model is a digital twin of the coast, as best CPRA can recreate, through correlation with hydrologic models and other studies of the ecosystem's response to things like water level, river flow, subsidence (sinking land, etc.) Those inputs come from the Coastwide Reference Monitoring System (CRMS) -- a large network of more than 300 wetland-based observation stations that, since 2006, have collected data about wetland elevation, water levels, salinity, vegetation, and land change. In the 2023 Coastal Master Plan, for example, data from CRMS stations has been used to refine how wetland vegetation in the ICM responds to changes in salinity and inundation stress.
  2. Storm surge and wave modeling (ADCIRC + SWAN) offers synthetic storm modeling, considering again the boundary conditions of river flow, water level, wind speeds, etc.
  3. The Coastal Louisiana Risk Assessment Model (CLARA) essentially models the flood depths and property footprint of the coasts to assess economic damages resulting from the interaction of the storm and the landscape model. Again, the model gets granular, assessing the first floor elevation of sub-communities and including two estimates of levee performance (no fragility, where levees hold; USACE study of fragility post-Katrina) and three estimates of pumping performance (fully operational; 50% capacity; no pumping).
  4. Each project's interaction with the landscape is independently modeled.

As should be obvious above, the model is highly reliant on the interaction between models, and in particular, the performance of the modeled projects as expected. Should any of them fail, or interact in unpredictable ways, the model's efficacy begins to decline. For example, in Barataria:

"As a result of the interactions of multiple diversions operating in a single basin, many of the additional diversion projects that had been effective as individual projects in building or maintaining land in the 2017 Coastal Master Plan showed mixed landscape results. The combined effects of the additional diversion projects and those already assumed to be operational resulted in excessive water levels in the basin. These results highlight the need for basin wide management of diversions."

This is true of any model, and the 6-year cycle of model updates recognizes the need to model and then verify. The task is immense, however, and the improved refinement of inputs and frequency of modeling runs might improve plan outcomes. That's like saying, "if someone wears glasses, they might see better", but it's true. As a corollary, any project too small to be modeled against the boundary conditions will appear as though it produces no resilience value.

The resolution of output achieved by ICM + ADCIRC/SWAN + CLARA. Also a glimpse at the immensity of the problem facing coastal Louisiana.

there is a modular architecture of funding for public projects

Despite the lack of federal funds, a number of projects have been built with state dollars to federal specifications. As a result, when federal dollars arrive (with their associated mandate of local contributions), the state and local entities can point to existing projects as their local match. While it's an elegant approach, it underscores the disparity in funding for poor states and rich states.

Typically, the federal government allows its funds to be spent before the local match, essentially giving state and local leaders years to raise the capital before they need to spend it. New York, for example, has spent more than $11B in federal funds on Superstorm Sandy recovery and resilience. They have $4B in additional federal funds to spend on projects that have been slow to deploy. Across more mature resilience projects with federal partnership, the city has spent less than $300M to date. That ignores some of the state's investment in grid hardening etc., but coastal resilience in other locales has been a largely federal initiative. As a result, there's been a decade of compounding to their benefit, while Louisiana continues to front the costs of climate change and energy production for the country.

voluntary acquisition is going to ramp up in the coming years

Communities have struggled with managed retreat and voluntary acquisition for decades. There are questions about who is relocated, to where, and for how much. The state of Louisiana has been running a number of these programs in concert with municipal and federal agencies. Some of them have not been going well.

In its master plan, the state recommends voluntary acquisition for every residence facing 14ft of storm surge or greater in a 1% AEP event. Those programs are going to get larger, as evidenced by the "best case scenario" in the plan below. Theoretically, the case below has less aggregate hurricane and storm risk than the state has today. This is partly due to the fact that many communities will have relocated.

Flood depths in 50 years assuming full program implementation and low climate change case.

Points I'm still exploring

what's the useful frequency and granularity of model inputs and linkages?

In other markets, we're exploring the benefit of bottoms-up capital formation as opposed to top down allocation, and I'm wondering if bottoms up model formation in concert with top down model design might have some interesting benefits to the resolution of the master plan model?

what are the real constraints on the master plan?

The state lists funding and sediment availability as critical bounds on the number of projects it can complete in a year. That's an insane statement until you consider the scale of the environmental engineering that is occurring. The models identify 41 different sources of sediment and scores them for renewability (all Mississippi River sources are renewable, all others are not). If federal dollars were to remove the cost constraint, how quickly would the plan run through available sediment?

A clear solution to that impasse would be sediment diversions, a sanitized way of saying, let the river run free and drop sediment in the coastal wetlands as opposed to the Gulf. Diversions, and in particular, diversions that undo the work of levees have been unpopular in the past for their impact on economic activity. My (limited) understanding is that the east side of the river is already getting crushed as changing wetlands cut into shrimping and oyster production, which may make diversion conversations less fraught.

will there be a reckoning for the cost of depleting coastal resilience?

In some ways there already is, as the people of Louisiana are bearing the capitalized cost of shaping the land. Collectively, the nation pays some ongoing opex too in the form of disaster relief funding each time storms exact an outsize toll on the state.

There are otherstakeholders, however, that seem to have avoided paying for their impact on resilience. The state's depletion of coastal land is unique in it severity -- largely brought on by diversions of the river for settlement as well as the hollowing out of intercoastal areas for oil and gas production. Thousands of small canals cut between wells and production facilities attacked the wetlands on two fronts by making them structurally susceptible to erosion, while changing the hydrology of the ecosystems which removed their ability to grow resilient vegetation.

There's no question that we've experienced collective benefit from these costs. For a long time, Louisiana's resources and the ports through which they traveled powered America. They still do, as much of the US' LNG exports run through Cheniere Plain. There's a severance tax on natural resources that are produced in LA for consumption out of state, but that's really the only socialized cost for lower resilience. What would it look like for there to be a resilience tax, in the same way that there's a bonding requirement for end of life oil wells? (The bonding requirement is a total joke, and is worthy of a post on its own, but a concept and its implementation are two different things.)

why is there no mention of power infrastructure or investment related to non-structural resilience?

In terms of decreasing economic costs of coastal resilience, grid hardening and other investments could prove massively accretive. I couldn't find anywhere in the master plan that these issues are discussed or explored.

land rights are crazy

There are so many exploratory questions around land rights -- how does sediment become land? Is that a legal distinction? Do you need a flag? When something becomes land, who owns it and which parts of it?

Relationship with private landownership was mentioned as an important point of focus in the master plan, and then not significantly discussed. Private land owners in Louisiana have significant rights relative to the rest of the country, and it can either slow or accelerate some of these projects.

The process for project selection.
A look at past plans, and the clear (projected) compounding of continued investment.
Sea level rise scenarios in the model.

Want to learn with us?

Sign up to get an email every time a new article is published. We share interviews, primary source data, books that provoked us, and reflections on emerging themes.