12.1. Gridded Data Files#

Model output has grown a lot bigger over the last decade and are becoming cumbersome to push those big files around to everyone. Instead, we can access the files remotely (meaning the files are on a computer somewhere in the world) and only bring back the small parts of what we need to do the analysis or draw the contours of the map we are making.

Current Model Output - Remote Access#

For current model output, the best remote site is:

https://thredds.ucar.edu/thredds/idd/forecastModels.html

Below are the URL filename structures for the most common synoptic models available through the Unidata THREDDS server. There are at least two weeks of current model output available via this access method. For the links below, replace the end of the URL (catalog.html) with the filename using the given format.

If you visit the main website above you’ll see that there are a whole lot of different model output sources, feel free to try and use them to make maps for tasks.

Current Model Output - Local Files#

Many institutions (or individuals) keep (download) a small set of local model output in netCDF (or other) format, which are easily read into Jupyter Notebooks. In a terminal window (or tab) locate where your netCDF model output is located.

Note

At Valparaiso University, local data from LDM can be found in the /data/ldmdata/ folder on our server. From there you can find were the model data are located using your terminal window, note the absolute path and the filename to be able to use in your code.

Record the absolute path to model data for each of the model types listed below:

NAM:

GFS:

All of the file names are of a standard format.

YYYYMMDDHH_<model>.nc

YYYY – 4-digit Year

MM – 2-digit Month

DD – 2-digit Day

HH – 2-digit Model Initialization Hour

<model> is the model name: gfs, nam (in lowercase)