Accurate weather forecasts are very important for sailors. Their predictions determine whether we set off on a passage, or wait for better weather. The information is used to pick a safe anchorage in which to shelter from a storm, or to identify a nice day to go for a cruise with friends.
We place a lot at risk on the information we are given, so why can some forecasts be so different, and why do we sometimes get completely different weather to what we were expecting?
Lets learn more about forecasting…
If you have been following my other Meteorology course posts, then you will already know that the weather we experience is a complex recipe of many different forces at work all around the world. Temperature, air pressure, seasons, the earth’s spin, ocean currents, moisture all mixing together.
With all these different weather ingredients mixing together in unequal measures, it is no wonder that the end product can be quite difficult to predict.
Forecasts use current weather observations from all over the globe and then run that collected data through computer generated models to estimate what is going to happen in the future.
The current weather is observed from weather stations on land, weather buoys at sea and satellites in space. This is standardised by the World Meteorological Organisation – a specialised agency of the United Nations to facilitate the global exchange of weather information. A good example of a local weather station is the Lyttelton Port Company weather page providing current and past weather statistics and a 48 hour forecast.
Weather stations provides hourly current weather information such as the temperature, dew point, wind speed and direction, cloud types and height, visibility, precipitation, swell/sea state, lightening and barometric pressure. Satellites provide a global view and Meteorological radar can provide information rain location and intensity. Upper air soundings from weather balloons are also used. Ships and planes can also provide weather data as they travel around the globe.
Once this data is collected, it is loaded in to the weather model as a starting point for a forecast. The model uses things called ‘primitive equations’ to determine rates of change in the atmosphere – this gives a prediction a short time in the future. The equations are used again on the predicted forecast to get another result for slightly further in to the future – and so on until you get a forecast for up to 7-10-12 days in to the future.
The charts that these models produce are called prognostic charts. This raw data is then usually combined with human interpretation and modifications taking in to account local influences, such as mountain ranges, or areas where wind is known to funnel and this is then produced as a forecast.
The trouble with the nature of these models is that weather dynamics are extremely chaotic in nature, and even a slight error in one of the weather observations can create vast anomalies in the mathematical models, particularly the further in to the future they attempt to predict. Therefore numerical models can only ever produce approximate solutions.
There are lots of different models – these are used by various organisations who then manipulate this data with their human influence or by running that through their own modelling system to produce a forecast. A couple of the models are as follows:
GFS – Global Forecast System
GFS is run by the United States National Weather Service. The data produced is free and is therefore used by a number of companies to produce weather forecasts. The model is run four times a day and produces forecasts up to 16 days ahead.
CMC – Canadian Meteorological Centre
The Canadian Meteorological Centre produces low resolution weather models for every 60 square km around the globe.
ECMWF – The European Centre for Medium Range Weather Forecasts
Based in the UK, this intergovernmental agency utilises supercomputer technology to produce numerical weather predictions to its European Member states.
These organisations produce the prognostic models which are then used by other organisations to produce the forecasts that we use.
Some of the different forecasts that are produced include:
- Synoptic Charts – showing large systems of high and low pressure showing air masses covering thousands of square miles. The picture above is a synoptic chart.
- Mesoscale – This scale deals with areas of only a few hundred square miles at a time – the size of a sea area. The forecasts give a reasonable indications of the average conditions of the area.
- Local Scale – This covers a smaller area over a shorter period.
Here are some of our favourite weather forecast providers:
PredictWind uses both GFS and CMC data and run that information through their own weather model to produce usable end forecasts specifically for sailors. They explain how they use the data provided by these weather models to create their own forecast:
The idea is that by running both models side by side you can gauge how accurate the forecast might be – if they are both saying the same thing, then you can be reasonably confident of the forecast. However if they are saying something completely different then you could assume that neither really know what is going on!
Here is an example of a prediction for NZ for a week out from today. Admittedly they were pretty similar up until this particular day:
PredictWind can provide all sorts of different forecasts including wind, waves, isobars, sea temperature, cloud, rain, satellite imagery and air temperature charts. They come in charts as above, tables and graphs. You can pay for extra features such as weather routing suggestions and you can also download an app to download forecasts while you are at sea.
What is a GRIB File?
GRIB stands for Gridded Information In Binary Form. GRIB’s are computer generated forecasts that can be sent in a compressed binary format – suitable to be transmitted via Satellite Phone or HF/SSB radio email systems.
Basically when you are out at sea and aren’t connected to fast internet, you don’t want to have to be downloading the highly graphical forecast content. Loading up websites can chew up lots of data, so GRIB files are the way to go and using PredictWind is one way of getting them. They combine an App with an Iridium Go satellite receiver and this has all sorts of other benefits (at a cost) too.
Metvuw is another online weather forecasting service that we like to use. They utilise the data provided by The United States National Weather Service – so the GFS model as mentioned above. That data is then run through a model written by James McGregor to produce the forecasts. There are forecasts for New Zealand and the South Pacific and also around the world. To read the charts you need to be able to read the wind feathers:
Here is the forecast for a similar time to the PredictWind forecast above. Lyttelton is located under the 998 isobar number – at 44ºS 0 the wind barb reads as a 15kt South Westerly, which is pretty similar to the PWG (or GFS) forecast on the left hand side of the Predict Wind chart. However there are definitely some differences in other areas on the chart.
Windyty uses the GFS model and also the ECMWF model. It is a non-commercial product created by a passionate wind sports addict called Ivo. It is mesmerising to watch the wind swirling around the high and low pressure zones. It gives a great overview of the whole world and you can zoom in and out to get a better perspective of your area. You can overlay the map with wind, waves, pressure, clouds and temperature.
I am not sure how useful this would be if you were at sea over a satellite internet connection, but if you are in an area with wi-fi or using internet data on your phone, then check it out: www.windyty.com
The Metservice is New Zealand’s weather forecasting agency. They are also contracted to do weather forecasting for other countries under the Metraweather banner.
Their weather forecasts are quite specific and there is different data available for the mountains, recreational boating areas, coastal marine forecasts, the Pacific High Seas, the Southern Ocean, and forecasts for towns and cities.
They also offer radio-fax services weather charts for the Pacific Ocean for sailors. They provide the information to Maritime Radio for their VHF radio weather reports which are read out regularly throughout the day.
There are literally hundreds of different weather forecasting services around the world. They all access similar data from those major super computer modelling systems to produce the local forecasts that we see every day. Accessing local forecasts – such as the Metservice here in New Zealand can be useful for understanding local anomalies.
Otherwise it is useful to subscribe to specialised services such as PredictWind for weather routing services and more marine specific applications. Get used to their different forecasts and information layers, learn how to download the data over satellite internet or SSB/HF radio services before you head offshore and away from the internet.
Listen to the weather broadcasts on the radio, and then use your other observations – reading the clouds, understanding the wind, monitoring the barometer readings, avoiding Tropical Revolving Storm seasons, utilising ocean currents in your favour, and being prepared for bad weather to hit at any time either when you are at sea or at anchor.
You can read my other Meteorological posts here:
- Atmospheric Pressure
- Atmospheric Moisture
- Ocean Currents
- Tropical Revolving Storms
- Reading a Barometer
If you have any favourite weather forecasting apps or websites, please mention them in the comments below 🙂