Home / Evidence and open data
Evidence
Elevation shapes physical risk, insurability and price. Here is the research we draw on, the lessons of recent events, the open data anyone can use to check a site, and the metrics we use to screen one.
Research findings
Markets already price some climate exposure, and miss much of the rest. These studies frame both the downside of low ground and the upside of the right high ground.
| Study | Market | What was measured | Finding |
|---|---|---|---|
| Bernstein, Gustafson & Lewis (2019) (opens in a new tab) Journal of Financial Economics | US coastal homes | Exposure to projected sea-level rise, comparing homes the same distance from the beach | ≈ −7% Larger discounts among sophisticated, non-owner-occupier buyers. |
| Keenan, Hill & Gumber (2018) (opens in a new tab) Environmental Research Letters | Miami-Dade single-family homes, 1971–2017 | Elevation and long-run price appreciation | Higher = faster Higher-elevation properties appreciated faster: evidence of “climate gentrification”. |
| Gourevitch et al. (2023) (opens in a new tab) Nature Climate Change | US housing stock | Parcel-level flood risk against current property values | $121–237 bn Overvaluation from unpriced flood risk, concentrated where disclosure is weak. |
| Benson, Hansen, Schwartz & Smersh (1998) (opens in a new tab) Journal of Real Estate Finance and Economics | Bellingham, Washington | Value of views, by quality and type | up to ~ +60% For unobstructed ocean views; partial and lower-quality views add far less. |
| Giglio, Maggiori, Rao, Stroebel & Weber (2021) (opens in a new tab) Review of Financial Studies | Very long housing leaseholds | Discount rates for cash flows a century or more away | Low rates Damage expected in 2080–2100 still carries real weight in today’s prices. |
| Befus et al. (2020) (opens in a new tab) Nature Climate Change; USGS summary | California coast | Response of shallow groundwater to sea-level rise | Flooding from below Rising seas lift groundwater inland, reaching basements, pipes and foundations behind the shoreline. |
| Rothermel (1972); Canadian FBP System (1992) (opens in a new tab) USDA Forest Service; Forestry Canada | Wildland fuels | Effect of slope on fire spread | ≈ 3× at 20° And about 6× at 30°, compared with flat ground. |
| IPCC AR6, Working Group I, chapter 9 (2021) (opens in a new tab) | Global | Sea-level projections | 0.44–0.77 m Median rise by 2100 (low to high emissions, vs 1995–2014). Close to 2 m by 2100 and 5 m by 2150 cannot be ruled out. |
Most valuation studies come from the United States, where transaction data is richest. Effects differ by market, disclosure rules and insurance systems; we treat them as benchmarks, not forecasts.
Recent lessons
Asheville, at roughly 650 m above sea level, was widely promoted as a climate haven. Helene’s rains devastated its valley floors and cut roads, power and water across the region. Absolute altitude is not protection; relative position and lifelines are.
Read: “there are no climate havens” (opens in a new tab)As nuisance flooding grew on the coast, investors turned to higher inland neighbourhoods such as Little Haiti. The resulting climate gentrification brought rising values, displacement and political scrutiny that long-horizon buyers need to factor in.
Read: climate gentrification and displacement (opens in a new tab)Dutch spatial policy now lets water and soil conditions steer where and how to build. On the elevated sandy soils that means retaining water and raising groundwater, limiting extraction near sensitive nature, and capping paved surfaces.
Explore the maps (PBL Atlas van de Regio) (opens in a new tab)Seawalls and dikes stop the sea at the surface, not underground. As sea level rises, coastal groundwater rises too and can emerge in basements, utility trenches and low streets well inland. Some barriers can even trap groundwater behind them.
Science and modelling (USGS) (opens in a new tab)Open data
The public datasets we draw on. They are independent sites with their own terms; we don’t see what you look up there.
| Source | Coverage | What it gives you | Use it for |
|---|---|---|---|
| IPCC AR6 Sea Level Projection Tool (NASA) (opens in a new tab) | Global coasts | Local sea-level projections to 2150, including vertical land motion | The sea-level horizon at a given coast |
| Copernicus DEM GLO-30 (opens in a new tab) | Global | 30 m surface model from the TanDEM-X radar mission (includes trees and buildings) | A first look at terrain and slope anywhere |
| OpenTopography (opens in a new tab) | Global, incl. European bare-earth model | Access to elevation datasets and processing tools | Bare-earth terrain where trees distort surface models |
| European Ground Motion Service (opens in a new tab) | Europe | Millimetre-scale subsidence and uplift from Sentinel-1 radar | Soft soils, subsidence and unstable slopes |
| AHN, Actueel Hoogtebestand Nederland (opens in a new tab) | Netherlands | 0.5 m LiDAR bare-earth and surface models | Parcel-scale height, basins and dike heights |
| Klimaateffectatlas (opens in a new tab) | Netherlands | Flood depth, cloudburst, drought and heat maps | Dutch multi-hazard screening |
| USGS 3D Elevation Program (opens in a new tab) | United States | LiDAR bare-earth elevation down to 1 m | US parcel elevation and slope |
| FEMA Flood Map Service Center (opens in a new tab) | United States | Regulatory 1% and 0.2% annual-chance flood zones | Insurance zones and 500-year checks |
| NOAA Sea Level Rise Viewer (opens in a new tab) | US coasts | Inundation scenarios and flood frequency | Visualising coastal exposure |
| Our Coast, Our Future (USGS CoSMoS) (opens in a new tab) | California coast | Storm flooding, cliff retreat and groundwater under sea-level rise | Coastal groundwater and erosion |
Metrics explained
Each metric can be derived from open elevation, hydrology and climate data. Together they replace “how high is it?” with “how well does it hold up?”.
Surface models measure trees and roofs, which can read many metres above the real ground in woodland. Screening always needs a bare-earth terrain model, or a wooded valley can look safe when it is not.
How far the ground sits above future extreme sea level. Behind defences, we also look at the residual depth if they fail and who funds them.
ground height − (local sea-level rise + storm tide + wave set-up)HAND is the decisive metric for river and cloudburst flooding inland. A mountain town can sit 650 m above the sea and 2 m above its river.
HAND = ground height − height of the nearest streamNegative values mark basins that collect water and cold or polluted air; positive values with a gentle slope mark water-shedding rises.
TPI = site height − mean height within radius RThe dry layer between the lowest floor and future high groundwater, allowing for sea-level rise near the coast and policy-driven water-table changes.
lowest floor − (high groundwater + sea-level effect + policy rise)About 1–8° drains by gravity without costly works. Above about 15°, landslide risk grows and fire accelerates uphill.
fire slope factor = exp(3.533 × tan(slope)^1.2)The lowest point, relative to flood level, along the route to the nearest main road, hospital and substation. Below zero, the site risks becoming an island.
L_min = min(ground − flood level) along the routeLine-of-sight rays from about 2 m above ground across a surface model measure the open view arc and the distance to visible water: the basis of view premiums.
Air cools by roughly 0.65 °C per 100 m on average. Basins trap heat and still air, while slopes and rises ventilate. See also Heat Protection Solutions (opens in a new tab).
Book a free 20-minute screen. We go through the sites or regions you are weighing, using these sources and metrics.