Follow one tide. Inspect the evidence.
Choose a peak tide, then play, pause or scrub through its rise and fall. The selected peak and the water level at the current playback time are different measurements. The map and playback bar identify the result actually displayed.
Saved peak-tide simulations
New simulations model bay-side tide entry only. Direct Pacific-side overtopping is excluded; water entering through bay-side breaches can still spread across connected land. Older saved results remain labeled with their original scope until replaced. Peak choices span 8′0″–9′6″ MLLW, one inch at a time. The default selection is 8′8″. Only completed saved simulations can open; choices marked preparing are unavailable. Selecting a height does not start a lengthy calculation in your browser. If the new default is preparing, the page identifies the older result it is displaying and opens the new default when it becomes available.
Each result follows an idealized 12-hour 25-minute cycle from 1′0″, up to its selected peak, and back to 1′0″, after two hours of low-water warmup. This is a hypothetical tide cycle, not a forecast for a calendar date. Results are calculated for that boundary curve, not resized from a different tide.
What the colors mean
Blue shows modeled water depth above bare-earth ground. Darker blue means deeper water. At overview zooms, narrow wet paths can appear wider; zoom in and click for the modeled cell values. Unshaded land is not proof of protection, and missing terrain is excluded from the calculation.
Red building outlines mean more than 20% of the complete mapped footprint overlaps modeled land cells with water at least about ¾″ deep. The percentage uses intersected area, not a count of cells. Exactly 20% does not trigger red. Unmodeled footprint area remains in the total area and is not assumed wet. Buildings excluded at your three requested locations are not highlighted.
Red does not confirm water indoors. Footprints come from OpenStreetMap; the model does not establish floor or door elevations, indoor entry, structural damage, or whether a building is impermeable. A building may stop being red as the modeled water recedes.
Water levels and vertical references
Water, wall and ground elevations in the main interface use one tide scale (MLLW), so they can be compared directly. Water depths are measured above the ground and distances are unchanged. Original survey values and references are explicitly labeled in source details. The solver and original data remain in NAVD88. The simulation uses the historic Newport Bay Entrance station 9410580 datum tie: MLLW is about 2⅛″ below NAVD88. The tie is from the 1983–2001 epoch; it is not a current local survey or a spatially varying conversion. Feet-and-inches formatting does not change the original stored values.
NOAA historical datum source ↗
How water moves
The hydraulic model conserves water and estimates surface flow, storage and bidirectional seawall overtopping. A solid barrier blocks that path until water rises above its crest; overtopping flow is reduced by water already behind it. All seawall openings are assumed sealed. Connected lower ground fills over time, and water can remain trapped after the tide falls.
Hydraulic cells are about 13′1½″ across, derived from terrain samples about 1′7⅝″ apart. Bare-earth cells do not represent finished floors or impermeable buildings. Uniform roughness and an assumed overtopping coefficient remain uncalibrated. Animation interpolates calculated snapshots spaced 2½ minutes apart; smooth playback is not finer physical resolution.
Location inspection reports ground and water depths in feet and inches, estimated speeds in feet per second, and arrival times. Shallow flow below about 2″ and estimated supercritical fronts have unresolved speeds. This display filter does not validate other speeds. The local-inertial approximation omits convective acceleration, jets and waves. Rainfall, storm drains, pumps, groundwater, infiltration, indoor flow and structural failure are excluded.
What the resolution check found
In a Peninsula test sector, reducing the hydraulic cell size from about 13′1½″ to 6′6¾″ shifted arrival times by 7½ minutes at the median and 35 minutes at the 95th percentile. Some cells switched between wet and dry. The largest local depth difference was about 7′10½″ at a shoreline sample on the elevation model’s flattened water surface, seaward of the modeled wall. The coarse-grid wall-proximity rule can incorrectly shield exterior water.
Excluding flattened-water samples still leaves material differences. These tests do not establish property-level depth, speed or arrival accuracy. The solver passes numerical benchmark and water-budget checks, but has not been calibrated against observed Newport flooding or established geographic mesh convergence.
View numerical sensitivity figures (PDF) ↗ · Download study results (JSON)
Terrain measurements
USGS CA_CaliforniaGaps_2_B23 was collected October 24–November 11, 2023. The native bare-earth terrain sample spacing is about 1′7⅝″. Horizontal reference: NAD83(2011), UTM zone 11N. Vertical reference: NAVD88 using GEOID18.
Project-wide DEM tests reported open-ground vertical RMSE of about 1¾″, a reported 95% value of about 3⅜″, and vegetated-terrain error of about 10½″ at the 95th percentile. These converted, rounded figures are not local street or seawall-top accuracy guarantees. Sample spacing is not vertical accuracy. Bare-earth processing can remove buildings and narrow wall crests.
Original terrain metadata ↗ · USGS tested accuracy results ↗ · USGS acceptance summary ↗
City walls and user corrections
The City’s July 24, 2018 completion report describes the Main Island north cap at 8′6″ NAVD88 and the south cap, up to Collins Bridge, at 9′0″ NAVD88. The August 2017 summary lists Grand Canal at 8′6″, Little Island east at about 8′8″, and Little Island south at 9′0″ NAVD88. These are rounded historical project records, not a current minimum-crest survey. Island cap heights are not applied to Peninsula walls.
City 2018 completion report ↗ · City 2017 height summary ↗
The City’s 600 January 2017 observations remain available as evidence. Only 528 explicitly labeled wall-top observations anchor survey-derived crests; bottom-wall, paving and control observations do not establish wall tops. SW wall-top geometry takes priority, with CW inventory as fallback. Heights interpolate between anchors; missing local crests inherit nearby published heights as qualified assumptions. The City point layer does not publish its vertical datum; treating its elevations as NAVD88 is a scenario assumption.
User-corrected heights and user-connected wall chains remain labeled separately from published measurements. Their original records are retained. The SW-1326 closure uses a user-selected crest, distinct from its recorded wall-base height. These corrections have not been independently validated. Missing linework or lidar returns are not proof of a breach.
We inspected original lidar returns and public engineering records, including American Legion record drawings. Their cap height varies, and an existing-grade callout is not a cap crest. Nearby lidar points or an approved design height do not establish the present minimum crest at a particular wall.
American Legion record drawings ↗ · City wall observations ↗
Building floor records
City elevation certificates are individually dated and may use NAVD88, NGVD29 or an unspecified datum. A footprint, adjacent ground elevation and finished-floor elevation describe different surfaces. Certificates may represent parcels or units rather than a uniquely matched building. Complete current floor or lowest-opening coverage has not been established; unknown heights remain unknown.
City elevation-certificate search ↗ · Certificate GIS records ↗
Map context and limits
Roads, buildings and coastline are from OpenStreetMap contributors (ODbL), extracted October 18, 2026. The coastline and low-elevation marine mask define the water boundary; positional errors can affect narrow shoreline features. Terrain is public domain; City GIS records retain the City’s data terms. Public elevation and engineering sources were reviewed in October 2026; a newer publication date does not establish newer terrain measurements.
This is an exploratory model, not a property-level flood forecast or an insurance determination. A defensible building assessment needs current surveyed wall crests and openings, floor and entry elevations, drainage and operational data, and comparison with observed floods. A smooth animation, unshaded property or one-inch selector does not establish protection or one-inch accuracy.