Using scenario based stress testing to discover medium (2050) and long run (2100) sea level rise risks

This project utilizes a scenario based qualitative stress testing approach to discover US coastal census tracts expected to adversely impacted by sea level rise (SLR) within the medium (2050) and long run (2100). One Baseline and two ‘plausible but severe’ Antagonistic scenarios were designed covering 1–7ft of sea level rise.

The important thing observations through this evaluation were that that outside of current FEMA high risk areas, there are ~3500 other census tracts along the contiguous US coastline which are expected to experience a mean SLR increase of 1ft within the medium term (by 2050). Long run (by 2100) a further ~1660 census tracts are in danger for two–7ft of SLR. The observations for the medium term are particularly essential on condition that they’re expected to materialize irrespective of future emissions pathways or other climate mitigation motion due the warming of the ocean that has already occurred as a consequence of climate change³.

This project predominantly utilizes Python data evaluation libraries, e.g., pandas and geopandas in addition to some visualization libraries, e.g., matplotlib, seaborn. The code for this project may be found here. The info for this project has predominantly been sourced from the US census bureau and from NOAA. Further details may be present in Technology and Data section below

Why Coastal Real Estate and Why Sea Level Rise?

About 40% of the US population, or ~128 million people, reside in coastal counties¹. While residential assets generally have been battered by a slew of maximum weather events lately, e.g., hurricanes, wildfires etc., for these coastal residents, one other significant emerging risk is flooding and property loss as a consequence of accelerated sea level rise (SLR). Nevertheless, within the exploratory research for this project, it was found that there may not yet be a deep understanding of the total extend of potential SLR impacts. Subsequently, this project intends to bring awareness to regions expected to be impacted within the medium term (2050) and long run (2100) in various scenarios of accelerated sea level rise.

The National Oceanic and Atmospheric Administration (NOAA) is a number one US scientific agency charged with monitoring oceanic and atmospheric conditions². Per NOAA’s 2022 Sea Level Rise Technical Report³, sea levels are currently rising at a significantly accelerated pace in comparison with previously observed levels within the twentieth century: sea levels along the contiguous US coastline are expected to rise, on average, as much over the subsequent 30 years (2020–2050) as they’ve during the last 100 years (1920–2020).

Over the long run (by 2100), SLR projections along the contiguous US coastline vary between 2–7ft on average in comparison with 2000 levels. The variability in these projections is principally as a consequence of uncertainity about future green house gas emission rates and resultant global warming, which is the important thing driver for SLR. Over the medium term (by 2050), given the warming of the ocean that has already occurred as a consequence of climate change³, contiguous US coastline sea levels are expected to rise between 1.3–2ft on average, irrespective of future emissions pathways or other climate mitigation motion. Moreover, these higher sea levels are also expected to cause more disruptive high tide flooding (HTF) events than occuring today: minor disruptive HTF events projected to extend from 3 events/12 months in 2020 to >10 events/12 months by 2050, and major destructive HTF events projected to extend from 0.04 events/12 months in 2020 to 0.2 events/12 months by 2050).

Project Goals

In light of those projections for coastal sea level rise for the contiguous US, the goals of this project are twofold:

1. To provide most of the people (potential home buyers, mortgage lenders, et al.) a tool to visualise the regional impacts of SLR over the medium term (2050) and long run (2100). Sea level rise is studied at various severities starting from 1ft to 7ft.
2. To offer specific information (FIPS codes) on the census tracts that may be impacted in each SLR severity scenario. Homebuyers and mortgage lenders can then consider this census tract level location information into their purchase decisions to access how much of a haircut, if any, they would love to use to their assessment of property value.

Scenario Based Approach to Assess Long Term SLR Impacts

Over the long run, given the potential variability in SLR projections, this project has been designed as a scenario based qualitative stress test. SLR scenarios have been designed based on the principles utilized within the Federal Reserve Capital Stress Testing exercise⁴; one baseline scenario and two “plausible but severe” scenarios of SLR have been utilized to evaluate how the varied coastal counties/census tracts throughout the contiguous US could also be impacted at various SLR levels over the long run.

The precise levels of SLR chosen for every of those scenarios were based on the 2022 NOAA SLR Technical Report³. Particularly, three NOAA SLR long run levels are considered on this project: Low (0.6m or approxiamtely 2ft), Intermediate (1.2m or approxiamtely 4ft) and High (2.2m or approxiamtely 7ft); these projected SLR levels have been mapped to the qualitative stress testing scenarios as shown below:

• A Baseline Scenario where sea levels are projected to rise 2 feet
• An Antagonistic Scenario where sea levels are projected to rise 4 feet
• A Severely Antagonistic Scenario where sea levels are projected to rise 7 feet

The Baseline scenario is mapped to NOAA’s long run “Low” SLR projection of roughly 2ft. This low projection is anticipated to materialize even at low emission levels with >92% certainity in any respect levels of world temperature rise3. The Antagonistic and Severely Advserse Scenarios are supposed to be “plausible but severe”. NOAA’s low term “Intermediate” SLR of roughly 4ft and “High” SLR of roughly 7ft, have been considered for the Advserse and Severely Antagonistic scenarios respectively. These high SLRs are related to higher green house gas emissions and corresponding larger increases in global temperatures in comparison with pre-industrial levels. Extreme scenarios of SLR greater than 2.5m (roughly 8ft) aren’t any longer considered plausible by 2100³ and so the utmost SLR level considered on this project is the “plausible but severe” NOAA High SLR of 7feet for the Severely Antagonistic Scenario.

Additional Evaluation for the Medium Term

Along with long run SLR impacts, this project also analyzes census tracts expected to be impacted by SLR over the medium term (2050). Over the medium term, SLR is sure on the lower end by 1ft and on the upper end by 2ft for the contiguous US. Recent scientific developments have led to an improved understanding of the aspects driving SLR and subsequently, there may be quite a bit more confidence on this narrower range (no matter the long run course of emission pathways)³. Attributable to this increase confidence within the SLR predictions over the medium term, this project doesn’t use a scenario based approach for the medium term. As an alternative, all contiguous US coastal census tracts in danger for SLR of 1ft by 2050 are considered in danger over the medium term on this project.

Evaluation of Current Risks vs Emerging Risks

This project analyzed 20 coastal states within the contiguous United States covering ~200 coastal counties and ~22K coastal census tracts inside them. Each census tract was analyzed by way of whether is it currently known to be in danger, i.e., “Current Risk Regions”, vs. if it may very well be in danger in the long run over the medium to long run, i.e., “Emerging Risk Regions”.

The primary a part of the project involved understanding current risks. When it comes to sea level rise, one in every of the best forms wherein it could present itself could be flooding. The Federal Emergency Management Agency (FEMA) has identified regions which are at high risk for flooding (FEMA floodplains)⁵ . This project used FEMA high risk floodplains as a proxy for areas which are currently known to be in danger for coastal flooding. On this project, these regions are considered “Current Risk Regions”.

The second a part of this project delved into identifying census tracts which may be in danger over the medium or long run. This project leveraged sea level rise information from NOAA to discover census tracts that would turn out to be an emerging risk over the medium term (2050) or long run (2100), i.e., “Emerging Risk Regions”. As a way to higher understand the severity of risk of rising sea leveles, emerging risk was divided into 4 categories — one category covering the medium term* and three scenario based long run categores:

• Emerging Medium Term Risk (1 ft SLR by 2050)
• Emerging Long Term Risk in Baseline Scenario (2ft SLR by 2100)
• Emerging Long Term Risk in Antagonistic Scenario (4ft SLR by 2100)
• Emerging Long Term Risk in Sev. Antagonistic Scenario (7ft SLR by 2100)

*as given there may be more confidence within the SLR predictions, no scenario based approach is required over the medium term.

Key Observations

The important thing observations through this evaluation were that that outside of current FEMA high risk areas, there are ~3500 other census tracts along the contiguous US coastline which are expected to experience a mean SLR increase of 1ft within the medium term (by 2050) and long run (by 2100) a further ~1660 census tracts are in danger for two–7ft of SLR.