Getting Started | Climate data

See this section in the introduction for some examples of Weather Routes using the CFS climate data.

There is a slight wrinkle when using the CFS forecast data in that the CFS model does not provide surface wind data. CFS provides wind at 1000mb along with the height of the 1000mb surface. In order to have the weather routing solver use this data, you must download both the 1000mb wind and height parameters.

CFS also provides ocean current data, which may be useful. Note that ocean currents are often small ocean features, and as the CFS data is only availble with a one degree resolution, many ocean currents are not represented.

Climate averages.

I know of a few Weather Routing systems which can operate on historic climate data. This is processed data, averaging the conditions at a point through a multi-year data set. This can also be useful when studing long routes, however, it does not provide the spread of possibilities offered when using CFS.

Averages tend to wash out extremes. When averaging, you tend to filter out maximums and minimums, moving all values toward the middle (by definition.) However, when performing pre-planning for a sailing voyage, its interesting to see some of the extremes that may be possible, the range of possibilities.

You can see this detail loss happening by studying the GEFS ensemble, and watching the mean (average) field and seeing how it relates to the individual members. The average field appears smoother, with no extreme highs or lows. Look into the individual members for the more extreme weather possibilities.

Nobody ever sails in a climate average weather system. Nobody will ever sail in any of the weather systems forecast by CFS either, however the CFS data does represent maximums and minimums, presenting you with more realistic challenges.

Historic data.

The best approach for the problem of studying the possibilities of long ocean routes may be to download actual weather data, from previous years. If you are studying a July departure for some event, if you could gain access to all July data for the previous 10 years, that might be useful.

For now, this remains as a possible future project for LuckGrib. Let me know if you would find this useful (and how much you would be willing to pay for it.)