LuckGrib does not make historic weather files available for download through the app. Only the most recent weather data is stored on the servers in the cluster.

In the past, I have had people ask for the ability to download historic data, and their needs tend to fall into two camps:

1. people who want to experiment with long ocean passages, and need enough weather data to sail across an ocean. The longest weather forecasts extend to 16 days, which may not be long enough for the Weather Routing system to cross an ocean in a cruising boat. For example, sailing from Californnia to Hawaii may take up to around 20 days in a typical cruising boat. The idea here is that by downloading historic data for several months, you will be able to experiment with longer ocean passages in the Weather Routing system.

2. people who want to study specific weather patterns which have occurred in the past, to study storms or study particular ocean races for example.

If you are in the first camp and want to experiment with long ocean passages, you may benefit from the CFS climate forecast data. For more details, please see:

• The CFS GRIB model
• Climate data in the routing system.

If you are in the second camp, and want to study specific weather systems from the past, continue reading.

There are potentially two methods you can obtain historic GRIB data. Each of these will require some software developer knowledge. The first method is a service run by Copernicus. The second is an archive of historic data run by the Iowa State University.

Copernicus historic data (Global.)

There is a service run by Copernicus which you may be able to use to download historic data. The link below leads to a web form hosted at Copernicus that provides access to historic GRIB data:

• Copernicus Climate Data Store.

You may need to register for a free account with that site in order for this process to work.

The form you end up on after clicking the link above allows you to download GRIB data from the ERA5 data set. This is essentially the initial few hours from every forecast run of the ECMWF, or Euro model.

Four types of data can be downloaded, you want reanalysis data. Reanalysis data is the most accurate record of the actual weather conditions that is known. As we all know, weather forecasts are pretty accurate for the first few days, and then fall off in accuracy over time. This reanalysis data is essentially the most accurate initial few hours of each forecast. This is a record of the global weather system capturing the state of the atmosphere which is later used in the forecast process. Reanalysis data is good.

You then need to choose some variables to download. To download 10m wind and sea level pressure, turn on the following variables:

1. 10m u-component of wind
2. 10m v-component of wind
3. mean sea level pressure

Then choose the year you want. If you want multiple years, you may want to download each year separately. Try to avoid downloading a single HUGE file, but prefer several smaller ones.

Then choose the months, days and hours you want.

If you want to download data for only a region, such as the Atlantic or Pacific, enter the lat / lon extents of the region. You can easily find these extents by scrolling the map in LuckGrib and noting the lat / lon’s.

Choose the GRIB format (not NetCDF.)

Iowa State University historic data (North America.)

The following link leads to an archive of weather data, in a variety of formats including GRIB data:

• Iowa Environmental Mesonet

Be sure to read the disclaimer from the link above.

Note that all of this data covers North America only, this is not global data.

This archive is organized by year, month, day, data type, organization, model and then parameters for each hour of the day.

To be really useful, you will need a way to concatenate many smaller files into some larger ones. For example, if you were to download all of the wind data for April of some year, you may end up with 60 files which need to be combined into a single one to be most useful.

Doing this may end up with very large files. You will want to do this work on a computer with a decent amount of memory.

To use this data source, you will need to be familiar with the Terminal app and some basic unix shell script commands. Or know a friend who is.

After downloading a bunch of GRIB files for the data you want, you can concatenate them all into a single larger file. This larger file is the one you will import into LuckGrib. For example, if you have downloaded the data into your Downloads folder, try something like this:

cd ~/Downloads
ls 202304*.grib2   # list of filenames for the downloaded data
cat 202304*.grib2 > 2023-04.grib2 # create a single large file

Then import the file 2023-04.grib2 into LuckGrib.

To download the data, visit the link above and first choose the year, then the month and day. Then:

• choose grib2
• choose ncep

There are three models, experiment with each of them. Depending on the model you choose, the remaining steps to obtain the data may be different. If you are using the RTMA data, wind is present in the UGRD / VGRD files, or WDIR / WIND. Use one pair of those or the other, not both.

If you go through this process and have any feedback that would help clarify this question, please pass it along.