Skill assessment of a total water level and coastal change forecast during the landfall of a hurricane
Why this work is in the frame
A frame that forgets how it found something cannot be audited. These are the routes that admitted this work.
Bibliographic record
Abstract
The Total Water Level and Coastal Change Forecast (TWL&CC Forecast) provides coastal communities with 6-day notice of potential elevated water levels and coastal change (i.e., dune erosion, overwash, or inundation) on sandy beaches that threatens safety, infrastructure, or resources. This continuously operating model provides hourly information for select regions along U.S. Gulf of Mexico and Atlantic Ocean coastlines. The objective of this work is to assess the skill of forecasts during a period of elevated water levels along the coasts of North Carolina (NC) and South Carolina, USA caused by Hurricane Isaias in August 2020, using a combination of observations and model hindcasts. Water levels and waves were observed throughout the storm at three locations near Wrightsville Beach, NC, which provided information to assess forecast skill; a wave buoy offshore, a tide gage at a local pier, and a pressure sensor deployed at the pier. In addition to observations, the non-hydrostatic phase-resolving model SWASH (Simulating WAves till SHore) was forced with hourly wave energy spectra derived from a coupled Delft3D-SWAN simulation during the peak of Isaias, to complement observations by computing nearshore wave height and wave-induced setup and runup at the shoreline. During the storm peak, SWASH-simulated water levels at the sensor position were comparable to those at the maximum landward extent (bias = −0.05 m; gain = 0.26; r 2 = 0.99), suggesting that observations at the USGS sensor location were a useful proxy for total water level (TWL; sum of tide, surge and wave runup) at the shoreline that are predicted by the TWL&CC Forecast. The TWL forecast at Wrightsville Beach was consistent with observations from the USGS sensor (bias = −0.38 m and −0.74 m, scatter index = 0.22 and 0.28 for the two forecast model grids considered, respectively; weighted regression considering model uncertainty explained 95 percent of variability in observed TWL). Observed TWL was within the confidence interval of the TWL&CC Forecast for the 5 h at the storm peak. Forecast mean water levels (MWL; sum of tide, surge and wave setup) and tide gage observations were also consistent (bias = 0.07 m and 0.02 m for the forecast model grids; scatter index = 0.46; r 2 = 0.80). Forecast MWL at the storm peak was within 0.06 m of the observed MWL from the tide gage for both sites. In the region where Isaias made landfall, eight additional pressure sensors were compared to the peak TWL forecast (bias = 0.14 m; scatter index = 0.18). Forecast TWL explained 90 percent of observed variability in TWL when considering uncertainty of the forecast with a weighted regression. The results demonstrate that wave-driven water levels contributed a significant portion of the forecast TWL during Isaias (52 percent during the three peak hours of the storm), and that TWL were represented using the forecast model. Mean absolute error of the coastal change forecast and observed overwash is 0.4 and 0.14 for the two forecast model grids considered. The skill demonstrated by this computationally efficient method indicates that the forecasting system can provide fast and reliable predictions of TWL across hundreds of km of coastline at sub-km resolution, days to hours in advance of when storms threaten coastal regions. • This is the first validation of the Total Water Level and Coastal Change Forecast. • Forecast total water levels compared to pressure sensor measured water levels. • Forecast total water levels compared to SWASH modeled water levels. • Wave-driven water levels contributed 52% to total water levels during the storm. • Regional forecast and observations of water levels and coastal change show skill.
Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.
Full frame distilled prediction
Teacher imitationNot calibrated prevalence, not ground truth. Human validation pending. Learned from the 10,348 direct Codex labels and 10,348 direct Gemma labels. Candidate is the union of thresholded teacher heads; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels or direct frontier model labels.
Codex and Gemma teacher scores by category
| Category | Codex | Gemma |
|---|---|---|
| Metaresearch | 0.000 | 0.000 |
| Meta-epidemiology (narrow) | 0.000 | 0.000 |
| Meta-epidemiology (broad) | 0.000 | 0.000 |
| Bibliometrics | 0.000 | 0.000 |
| Science and technology studies | 0.000 | 0.000 |
| Scholarly communication | 0.000 | 0.000 |
| Open science | 0.000 | 0.000 |
| Research integrity | 0.000 | 0.000 |
| Insufficient payload (model declined to judge) | 0.000 | 0.000 |
Machine scores (provisional)
The two teacher heads of the student model, read on this work. A score orders the frame for review; it never asserts a category, and the validation status ships verbatim with every row.
Baseline scores from an immature model (maturity gate not passed, 7 training rounds). Scores rank; they never assert a category.
score_only:v0-immature-baseline · verbatim from the scoring run: score_only means the number may rank works, and no category label ships from it