How NBFlood Measures and Calibrates Flood Risk

In short: We use the UC Davis gauge because its continuous readings broadly fit what we observed at a surveyed seawall and reveal changes a single tide alert can miss. Separately, NBFlood’s Observed scenario raises modeled ground by six inches to bring street flooding closer to what we observed when comparing the measured tide, surveyed wall height, and separate lidar elevations. The tide and wall heights do not get that adjustment.

At a seawall, a few inches can decide whether water stays in the harbor. To judge that gap, we need to put tide readings and wall heights on the same scale.

Tide tables and live gauges

A tide table is a prediction of the astronomical rise and fall. NBFlood uses NOAA’s Newport Bay Entrance prediction. It does not include every change caused by weather and ocean conditions. NOAA’s station record says its entrance gauge was removed in 1994, so this prediction is not a live reading from that gauge.

A gauge measures water that is actually there. The UC Davis Newport Bay station, installed with the City’s Harbor Department, monitors the bay around the clock when operating. Its dashboard reports water relative to NAVD88. NBFlood displays five-minute averages on the same Mean Lower Low Water (MLLW) scale as the NOAA prediction; missing readings stay missing. Gauge metadata and NBFlood’s conversion method.

The City also operates a Balboa Yacht Basin gauge. It records water every six minutes and transmits provisional readings hourly. In matched readings from October 23rd, UC Davis averaged 5.43 inches higher. That sample does not establish a permanent offset or prove either gauge wrong: they are at different sites, and we have not independently checked their survey ties.

A publicly announced high tide is another kind of number: one reported value without the time history. Continuous readings show when water rose, how long it stayed high, and whether a shorter rise occurred before a selected high-tide reading.

Where the data comes from

The tide, wall, and terrain records come from government agencies and UC Davis, a public university. The visual wall clearance and six-inch ground adjustment are NBFlood observations and modeling choices.

Checking water against the wall

On October 27th, the City announced a high tide of 8.4 feet. That is decimal feet, or about 8 feet 5 inches. The announcement does not name a gauge, measurement time, or datum. It cannot, by itself, tell us how close water came to a particular wall.

We watched water near the City’s SW-1221 wall point, whose recorded top is 8.61 feet. Because the City’s Balboa Island seawall report uses NAVD88, NBFlood provisionally interprets SW-1221’s height on that datum too. Adding NOAA’s local datum difference of about 0.18 foot puts the wall near 8′9½″ MLLW for comparison with the tide. The SW-1221 layer does not confirm its own datum, and the entrance-station tie may differ slightly at the wall.

We estimated water about three inches below the wall top. Around the UC Davis peak, 13 five-minute readings averaged 8′7.3″ MLLW over one hour. On our provisional scale, that is about 2.2 inches below the wall—close to our visual estimate. If SW-1221’s 8.61 feet were already MLLW, the same gauge mean would be nearly level with its top. This comparison supports our interpretation; it does not prove it. Waves, wakes, and local differences can change the water at the wall. UC Davis archive; NOAA datum sheet.

The one-hour UC Davis mean and NBFlood’s provisional MLLW conversion of SW-1221’s recorded 8.61-foot crest leave about 2.2 inches, broadly consistent with the roughly three-inch gap we saw. This is an illustration, not a photograph or surveyed section; the City’s SW-1221 layer does not specify its datum.

The City’s 8.4-foot announcement would imply about 4.7 inches below our provisionally converted wall top if it used the same MLLW reference. That is farther from what we saw, but the announcement and our observation are not matched readings. We cannot call the City’s number a mistake or assume it came from the Balboa gauge.

The UC Davis record shows both the short peak and the surrounding average. The City’s announcement is a separate callout because its measurement time and source are unstated.

The full UC Davis record is useful even if our wall estimate is off. It shows short peaks and sustained levels; one announced value cannot. In our limited gauge comparison, UC Davis also read higher than the City gauge. Using that higher reading as a cautious screening reference leaves less apparent clearance below walls or sandbags. It does not mean UC Davis is always higher or more accurate at every barrier.

November 1st showed why timing matters

From 2:30 to 5:30 a.m. PST on November 1st, UC Davis water averaged 28.4 inches above NOAA’s prediction. Eight readings reached at least 30 inches above it.

Across the day, water averaged 25.3 inches above prediction. The morning was about 3.1 inches above that already elevated daily average. The chart shows a sustained gap from the tide table, rather than a sudden 28-inch jump confined to those three hours. Archived gauge readings, NOAA predictions, and audited daily average.

The pale line shows five-minute UC Davis readings above the NOAA prediction; the dark line smooths them over 30 minutes. The dotted line is that day’s average excess, a retrospective reference. The vertical scale is zoomed in and does not start at zero.

A single reading at a chosen high-tide time could miss this earlier interval. It might also miss how long water stayed near a low wall. The City’s Balboa gauge takes repeated readings too; the information lost is in a single announcement or manual observation, not in the City’s instrument itself.

A five-inch temporary rise can turn four inches of clearance into one inch of overtopping. This is an illustrative clearance, not a measured sandbag height or a reconstruction of November 1st at a specific barrier. The City offers sandbags, but we do not have measured crest heights for individual installations.

Heavy rain in the region may be relevant. An archived National Weather Service advisory preceded the morning readings and included Newport Beach among areas at risk. We have not established that rainfall caused the water-level excess. We want to compare local rain, runoff, wind, pressure, and gauge records on one timeline before drawing that conclusion.

Why we raised modeled ground six inches

Water height alone does not tell us where flooding will go. NBFlood starts with USGS bare-earth lidar in NAVD88. We use NOAA’s published datum tie to compare it with MLLW water. That conversion is about 2.2 inches and is documented in our terrain audit.

The datum conversion alone did not match our observations along low walls and streets. Raising the modeled ground by six inches brought the simulated street waterline closer to what we saw. This is an empirical scenario. We are not saying the USGS elevations are wrong; they do not appear to line up with the measured tide against the seawall in this simulation. We still need dated, location-specific photos or surveys tied to gauge readings to reproduce that calibration. The wall audit and scenario definitions describe the current evidence and implementation.

In NBFlood’s Observed scenario, ground rises six inches. Tide forcing and wall crests stay put. Reported retains the original ground.

For example, a modeled ground cell shown as 9′0″ in Reported appears as 9′6″ in Observed. Our observations suggest a similar apparent difference may apply when comparing a home’s surveyed elevation with the UC Davis tide scale: a surveyed 9′0″ might look closer to 9′6″ in that comparison. This is a separate, unverified observation, not a property-specific correction. A floor, doorstep, adjacent ground, and lidar cell are different features. For decisions about protecting your home, rely on its official elevation certificate or survey, and confirm its datum and the feature measured.

Some elevation certificates are publicly available through the City, but they do not cover every home. We lack a reliable ground-floor or lowest-opening height for most properties, so NBFlood cannot say when water would enter a particular home. The Flood Sim’s address search shows the deepest modeled water depth at and immediately around a building during the selected tide cycle. You can use that estimate with your own floor or doorstep survey and what you know about your property to assess when it may be at risk. It is not a prediction of indoor flooding. Floor-elevation audit.

Help measure what protects the shore

Water and protection are moving targets. Gauges record changing conditions; wall surveys and lidar describe conditions when they were collected. Temporary sandbags, berms, pumps, and tide valves can change the route water takes. A useful forecast needs to know what was in place when water reached it.

We want to record each barrier’s location, lowest crest height, height reference, installation date, and any openings or overtopping. The City’s public alerts describe protective work, but the public records we found do not give a dated, location-by-location set of barrier crest elevations. If the City has these measurements, sharing them would help us compare protection with the gauge record.

Residents, harbor workers, surveyors, and City staff can help. A dated photo, clear location, and note about whether water reached or crossed a barrier are useful; a surveyed height and datum are even better. Please leave protective work undisturbed, as the City advises. Message @nbflood on Instagram or contact Chad Huntley on LinkedIn. We can place those observations beside the measured tide and improve the next forecast together.

Illustration, not a documented City installation. We still need measured heights and dates for actual protections.