How WEMo compares
WEMo began as NOAA's Wave Exposure Model in 2007. This version keeps the original as a setting and adds the shallow-water physics, global data and web workflow it lacked. Here it is side by side with NOAA's WEMo, other fetch-based tools and SWAN, the spectral wave model it was tested against.
Side by side
"Optional" marks an add-on you switch on for a run; "Partly" marks something limited or left to you. Scroll sideways on a small screen.
| Capability | NOAA WEMoArcGIS 9.x (2007–) and the newer R package | Other fetch-based toolsUSGS Wind Fetch and Wave Models, REI indices, R fetch packages | WEMo by Geo HorizonsThis web app (WEMo 2) | SWANThird-generation spectral wave model |
|---|---|---|---|---|
| Getting started | ||||
| Type of model | Fetch-based; the 2007 formulation | Fetch lengths, an exposure index, or wave estimates from fetch and wind | Fetch-based, with shallow-water physics tuned against SWAN; the 2007 formulation kept as a setting | Solves the wave energy balance on a computational grid |
| Where it runs | ArcGIS 9.3 on Windows (version 4.0), or R | ArcGIS toolbox (USGS) or R | YesAny web browser; the model runs on our servers | Command-line program on your computer or a cluster |
| Software and cost | PartlyFree tool; the ArcGIS version needs an ArcGIS licence (ArcInfo/ArcEditor for some shoreline steps); R is free | PartlyFree tools; the USGS toolbox needs ArcGIS with Spatial Analyst | YesNothing to install; free during early access | YesFree, open source; you add tools to prepare inputs and view results |
| Data you prepare | NoShoreline polygon, depth grid and wind records (the R version adds helpers for NOAA data) | NoShoreline or water polygon and wind; depths for the wave tools | YesNone: NOAA Coastal Relief Model depths on the US coasts and GEBCO 2026 elsewhere, hourly wind back to 1940 and nearby buoys are built in; upload your own if you have them | NoGrid, bathymetry, wind fields and boundary conditions |
| Skills needed | GIS (and R for the R version) | GIS or R | YesDraw an area on a map | Wave-modelling expertise |
| Wind and depths | ||||
| Wind | PartlyA station record (NDBC format) or speeds and frequencies for eight headings | PartlyWind you supply | YesHourly reanalysis over water, blended between cells; or a nearby buoy or station from four archives; or your record. Any period back to 1940 | PartlyWind fields you supply, steady or changing in time |
| Depths | PartlyYour depth grid | PartlyYours, where used | YesNOAA Coastal Relief Model (about 30 m) on the US coasts, GEBCO 2026 worldwide, or your survey blended in; tide, storm or lake water levels | PartlyYour grid |
| Physics | ||||
| Directions | Partly16–56, one straight ray each | PartlyFixed radials | YesUp to 56, three rays each | YesFull directional spectrum |
| Growth in shallow water | NoDeep-water curve everywhere: heights keep growing with fetch | PartlyEmpirical growth curves | YesStops at the depth-limited height (Young and Verhagen) | Yes |
| Bottom friction | No | PartlyVaries | YesJONSWAP, as in SWAN | Yes |
| Breaking | PartlyFixed bed slope | PartlyVaries | YesLocal bed slope along each ray | Yes |
| Energy from neighbouring directions | NoNo: sheltered sites read too low | PartlySome indices average nearby radials | YesDirectional spreading | Yes |
| Refraction | No | No | PartlySwell always; wind waves with detailed (≤100 m) bathymetry | Yes |
| Ocean swell | No | No | OptionalCarried into bays, sounds and straits | YesYes, from boundary conditions |
| Vegetation | No | No | OptionalSeagrass or marsh areas you draw | Yes |
| Coastal structures | No | No | OptionalSeawalls, revetments, breakwaters, sills: sheltering, transmission, reflection, toe scour | YesObstacles with transmission and reflection |
| Currents, storms in time, wave setup | No | No | NoNo: results summarise the wind climate | Yes |
| Results | ||||
| What you get for each site | PartlyWave energy (RWE), height and period; optional bottom velocity and sediment motion | PartlyFetch or an index; the USGS tool adds wave height, period, orbital velocity and shear stress | YesWave height, period, RWE, orbital velocity, bed shear stress, five exposure classes | YesWave height, periods, direction, spectra and more, on the grid |
| Maps, reports and files | PartlyGIS layers in ArcGIS, or R objects and plots | PartlyGIS layers or tables | YesInteractive map, PDF report, CSV, GeoJSON and GeoTIFF; projects saved online | PartlyOutput files for your own processing |
| Hourly time series | PartlyA wind climate or a chosen wind per run; no hourly series | PartlyWaves for a chosen wind; series by scripting | YesHour by hour at up to 25 sites, with duration-limited growth, compared with NOAA buoys or your measurements | YesNon-stationary runs, with your own wind fields and boundaries |
| Scripted access (API) | PartlyThe R version can be scripted; the ArcGIS version runs by hand | PartlyR packages can be scripted | YesWeb API with personal keys: run from Python, R or curl and get JSON, CSV, GeoJSON or GeoTIFF back | PartlyScripted runs on your computer, with your own input preparation |
| Accuracy and speed | ||||
| Agreement with SWAN (939 lower Chesapeake Bay sites, 16 steady winds) | Partly71% of sites within 20% at 10 m/s, 45% at 20 m/s; error 0.13 m and 0.43 m | Not tested here | Yes98% within 20% at 10 m/s, 84% at 20 m/s; error 0.05 m and 0.28 m | Reference |
| Checked against measurements | YesValidated in the 2007 NOAA memorandum | PartlyVaries by tool | Yes31 wave buoys, hour by hour, 2022–2024, in four basins | YesWidely validated |
| Speed and scale | PartlyDesktop; NOAA calls it CPU-intensive (the R version is faster) | PartlyDesktop | YesAbout 1,000 sites in 10 s and 16,000 in under 2 minutes; up to 50,000 sites a run, many users at once | NoA grid run for each wind condition; a long-term wind climate needs many runs |
Which to use
WEMo fits when you need to…
- Compare wave exposure across many sites in a bay, estuary, sound or lake.
- Screen sites for seagrass, oyster reefs, living shorelines or aquaculture.
- Summarise years of wind into one map and report, quickly and consistently.
- Test a seawall, breakwater or sill at the screening stage.
Use SWAN or another detailed model when you need…
- Engineering design conditions, with a coastal engineer.
- Waves through a particular storm, hour by hour.
- Currents, wave setup or diffraction around structures.
- Open-coast wave fields where swell dominates.
Notes and sources
Agreement with SWAN: both models run on identical depths and 16 steady winds (10 and 20 m/s from eight directions) over lower Chesapeake Bay; 3,677 site–wind pairs per speed. The 2007 column uses the original settings with one straight ray per direction. Over the whole bay, ten times the area and not used to set WEMo's defaults, 97% of 24,586 site–wind pairs were within 20% of SWAN at 10 m/s.
Speed: default settings (56 directions, three rays each, 100 km fetch), wind already downloaded, on a 2-thread test machine; a run in a new area first downloads its wind. Expect two to three times longer on an older server.
NOAA WEMo details are from the WEMo 4.0 manual and NOAA's WEMo pages; other tools from their own documentation. Sources: Malhotra and Fonseca (2007), NOAA Technical Memorandum NOS NCCOS 65; Booij et al. (1999), J. Geophys. Res. 104(C4); Malhotra (in preparation), WEMo 2 and the WEMo–SWAN–buoy benchmark. See also Why WEMo, How it works and its limitations and the guide to wave exposure tools.
WEMo on the web is an independent product of Geo Horizons, not affiliated with or endorsed by NOAA. SWAN is developed at Delft University of Technology.