If you have ever checked a surf forecast to see if a weekend swell is hitting your favorite break, you have probably relied on a powerful tool called WaveWatch.

The computer model is one of the most widely used for predicting how waves travel across the ocean.

The story began in the late 1980s at the Delft University of Technology (TU Delft) in the Netherlands.

A scientist named Hendrik L. Tolman started developing a new way to track how wind transfers energy into water, as part of his thesis work.

The first version, known as WaveWatch I, focused on the basic math of how waves grow and move.

“The development of WaveWatch I was entirely funded through my Ph.D. work at Delft University,” Tolman once told.

In other words, WaveWatch I was a one-man (and one-university) show.

The early version ran on large mainframe computers and was already a spectral model, meaning it tracked the energy of waves across many lengths and directions.

It was a significant step beyond the early forecasts of the 1950s and 1960s, which were largely empirical or very simple and, obviously, a massive change from World War II’s very simplistic predictions.

Early 1990s: Improved Equations

After finishing his Ph.D., Tolman took his idea to the United States.

While working at the NASA Goddard Space Flight Center, the scientist developed WaveWatch II, a major step up from what had been initiated in the Netherlands.

It began to use more complex equations to account for how waves change as they move through different water depths.

Under NASA’s roof, WaveWatch II was rewritten and improved. It ran on NASA supercomputers and helped with ocean missions.

The model’s physics and numerics got better, and it could use NASA’s atmospheric data.

Although only one research team was working on it, WaveWatch II demonstrated that the model was ready for wider use, namely, daily public forecasts.

The Third Iteration

By the late 1990s, the U.S. National Oceanic and Atmospheric Administration (NOAA) was ready for a new wave model.

The old NOAA wave model was based on WAM, an earlier third-generation model from the 1980s.

NOAA saw that Tolman’s WaveWatch had potential to become the next big thing.

They funded the creation of WaveWatch III and pulled Tolman (and his code) into their Marine Modeling and Analysis Branch (MMAB) at the National Centers for Environmental Prediction (NCEP).

Unlike the older versions, WW3 doesn’t just guess how waves will look based on general patterns.

Instead, it calculates the energy of the waves at every single point on a grid across the entire ocean.

It looks at three main things:

• How the wind builds the waves up;
• How waves interact with each other and lose energy;
• How the ocean floor and obstacles like islands slow them down;

Hendrik Tolman described the complexity of this process by noting that “waves at sea represent a stochastic process, and that a single representative wave height does not adequately describe the wave field at sea.”

In simpler terms, the ocean is too messy for simple math, so WW3 uses spectral descriptions to track many different wave groups at once.

The first publicly available version, around 1999, was coded in modern Fortran and set up to run on NOAA’s new supercomputers.

In fact, the model was fully implemented and turned operational in March 2000.

So, on March 9, 2000, NOAA switched its ship and buoy forecasts from the old WAM-based system to WaveWatch III.

In other words, NOAA made WaveWatch III its official global wave model at the turn of the millennium.

Open Source and Mergeable

WaveWatch III was a significant upgrade.

It solves the full wave energy spectrum with physics – wind growth, wave breaking, currents, ice, and more.

It runs four times a day, matching the main weather model cycles, and produces maps of significant wave height, period, and direction across the globe.

They’re actually publicly available.

In fact, in 2019, NOAA moved the code to GitHub and invited the whole community to help.

Hundreds of scientists and forecasters from the U.S., Europe, Asia, Australia, and beyond now use and improve WaveWatch III.

The code base is open source and constantly updated with the latest science.

In recent years, the team even added shallow-water surf zone options and ways to couple waves to storm surge models, so WaveWatch can reach coastal forecasts, not just the open ocean.

Anyone can witness the visual and graphic representations of it, for instance, displaying large groundswells steadily crossing oceans.

All you need to do is go online to NOAA’s Environmental Modeling Center (EMC) and check the forecasts for the following products:

• Significant wave height;
• Peak wave period;
• Wind sea wave height;
• Wind sea period;
• Wind speed and direction;
• Primary swell wave height;
• Primary swell period;
• Secondary swell wave height;
• Secondary swell period;

Today, WaveWatch III is incredibly precise and has been merged with several atmospheric/weather models, including the Global Forecast System (GFS) and the Global Ensemble Forecast System (GEFS).

It even includes special “tricks,” like handling “unresolved islands,” which are tiny specks of land that used to be too small for computers to see but still blocked swell, and reducing the “garden sprinkler” effect, an artifact that can make waves spread unrealistically.

WaveWatch III Has a Wide Use Globally

Today, WaveWatch III is the engine behind almost every surf report you read.

When sites like Surfline or Stormsurf show a long-period swell traveling from a storm near Antarctica toward Hawaii or California, they are likely using data from NCEP’s global implementation of the model.

The U.S. Navy also prefers WaveWatch III for open-ocean wave prediction.

Coastal observing systems in Hawaii, Puerto Rico, the Gulf of Mexico, and elsewhere display local WaveWatch maps on their websites.

There are even smartphone apps that tap into NOAA’s WaveWatch/GFS/GEFS data to provide surf predictions.

It is also important to note that forecasters compare WaveWatch outputs with buoy measurements to fine-tune local reports, for instance, at the regional level.

The model is still embedded in several proprietary systems developed privately by companies, national organizations, and weather services.

Words by Luís MP | Founder of SurferToday.com