Municipal exposure to sea-level-rise inundation in Connecticut: a reproducible screening analysis of NOAA inundation-extent data and its sensitivity to methodological choices

Arav Siddharth Mestry 

Municipal tabulations of sea-level-rise exposure are widely used in coastal planning, but they are sensitive to processing choices that are rarely reported. We tabulated NOAA Office for Coastal Management ocean-connected inundation-extent rasters for Connecticut (0.5–10 ft above mean higher high water, MHHW; ~3 m resolution from 2023 statewide lidar) for all 169 municipalities, using 2020 Census boundaries, exact ellipsoidal cell areas and the present-day MHHW extent as a baseline mask. Newly inundated area is 34.7 km² at 1 ft, 51.3 km² at 2 ft, 68.5 km² at 3 ft and 133.3 km² at 6 ft, affecting 45 to 52 municipalities. About 70–75% of this area lies in the 24 Long Island Sound shoreline municipalities, and Old Saybrook is the most exposed in proportion to its land area (9.8% at 1 ft; 18.7% at 6 ft). Land cover shows that most newly inundated area at 1–2 ft is existing tidal marsh (developed land is 3–6%), whereas developed land reaches 27% at 6 ft. Exposure also extends up the tidal Connecticut River beyond the statutory coastal area, a signal that depends on NOAA's inland extrapolation of the tidal surface. Municipal rankings are conditional on scenario and metric: between 1 and 6 ft, Spearman's ρ is 0.74 for absolute area and 0.69 for share of land, and only 6 and 4 of the respective top-ten municipalities are shared. Omitting the MHHW baseline mask overstates exposure 4.6–14.7-fold, whereas generalized boundaries and aggregation to ≈81 m change statewide totals by less than 10% and leave rankings nearly unchanged. Low-increment estimates are sensitive to small vertical offsets (−40% to +27% at 1 ft for ±0.5 ft). A coordinate-based heuristic in an earlier dashboard by the author correlated weakly with these estimates (ρ = 0.45–0.49) and should not be used. The results are screening-level estimates of potential permanent high-tide inundation, without independent validation, and are released with a fully reproducible workflow.