Weather Forecasting - Which Forecast To Use

Weather forecasting is big business, and it matters enormously for recreational boaties, since weather can be the difference between an idyllic adventure and a bone-rattling misadventure. Forecasts are built from computer weather models that carve the atmosphere into a 3D grid, with accuracy improving as grid size shrinks and computing power increases. Forecasts are most reliable in the first few days and lose accuracy rapidly beyond about seven days. For the specific items to check before heading out, see our guide on boating safety: check the weather forecast, which steps through marine warnings, wind, waves and tides in order.

How Are Weather Forecasts Made?

Computer weather models, also known as numerical weather prediction, carve up the atmosphere into a 3D grid and use algorithms to predict future weather. The grid size is critical to accuracy and is shrinking as computing power increases. The current global grid size is 25km, but this is reducing to 12km with local enhancements down to 1.5km, and local grids of 100m are expected from supercomputers such as the UK Met Office’s, which cost A$2.2 billion to acquire from Microsoft in April 2021.

All weather models are derived from pretty much the same raw data. A huge range of physical observations, including air pressure, humidity, wind and temperature at various levels above and below sea level, are collected globally and made available to meteorological agencies. These agencies then run their computer models and sell or distribute the output to national and commercial weather services. The raw data is typically already around six hours old before the forecast process even begins, once you account for the time taken to collect it and run the algorithms.

COVID 19's Impact on Global Weather Forecasting

When flights around the world were grounded at the height of the Covid-19 pandemic in 2020, weather forecasts became slightly but materially less accurate. Aircraft are the second most important source of atmospheric data after satellites, since they continuously collect wind and temperature readings at altitude, and the loss of up to 75% of those observations created real gaps in the models.

However, the story didn’t end there. When planes returned to the skies, the sudden surge in aircraft data volumes actually created new problems. The European Centre for Medium-Range Weather Forecasts (ECMWF), one of the world’s leading meteorological bodies, suspended a type of aircraft observation known as Mode-S data in November 2022 because the sheer volume of returning data was overwhelming its data assimilation systems and paradoxically degrading forecast quality. By January 2024, ECMWF had reintroduced this data using a new ‘box thinning’ approach that uses one aircraft report per atmospheric box rather than all available readings. Short-range upper tropospheric wind predictions over Europe are now up to 8% better as a result.

In March 2025, the UK Met Office formally adopted a similar approach, partnering with Flightradar24 and the European Meteorological Aircraft Derived Data Centre to integrate data from thousands of aircraft daily into its global model. The aircraft data issue created by the pandemic is well resolved, and in several respects forecast accuracy has improved beyond its pre-2020 baseline as a result of the better systems now in place.

Local Enhancements Matter

Once you have the raw global model, local enhancements are made. For example, Perth’s famous sea breeze is always underdone in the global model, so the Bureau of Meteorology tweaks the model to take into account local effects throughout Australia. The MetEye website is a strong choice for this reason. Seabreeze is another popular Australian site among recreational boaties, having provided coastal wind and swell forecasts for over 25 years, and its easy-to-read format suits those who want a quick, location-specific read before heading out.

How Far Ahead Can I Trust a Forecast?

Forecasts made by weather models are most accurate for the first few days and rapidly lose accuracy at anything longer than 7 days. This reflects the chaotic nature of weather, in which exceedingly small uncertainties in the current state of the atmosphere have a mushrooming influence on forecasts at longer time ranges.

One technique to test the confidence of forecasts is to run the model multiple times with slightly different starting conditions or times. This tells you whether small changes in the atmosphere or ocean are likely to cause a significant shift in the forecast.

Use of Forecasts

Most people have their favourite weather website based on usability and perception of accuracy. Whatever site you use, it’s worth also using the PredictWind app to get a feel for the overall certainty of the forecast. PredictWind, now trusted by over one million users worldwide, shows the results of multiple models simultaneously, including its own proprietary AI-powered models as well as ECMWF, GFS, UKMO and others. If all the models align, you can be fairly confident in the forecast; where there are large variations between the different models, you need to be more careful.

Windy is another visually intuitive option that displays forecast wind and swell data on an animated map, and is particularly useful for planning longer coastal passages where conditions vary across different sections of the route. And one final thing: a look out the window usually provides the best short-term forecast.

Frequently Asked Questions