and I'm very interested in trying to get this to work with New Zealand data. I'm just wondering in the 'Download Radar Data' python script, how this georeferences the raw images? and would it be possible to change this script to download New Zealand data?
The Python script gets two files from the National Weather Service's website: an image file in the .gif format and a .gfw file (AKA world file). The gfw file has spatial reference information that is used to georeference the radar image.
One way the NWS distributes its radar data is through the following website: http://radar.weather.gov/ridge/Conus/RadarImg. You can click on one of the links and save the data from that link to your local drive. The Python script just imitates this process; the specific files that are downloaded depend on what the user specifies in the parameters ("northeast sector", "southeast sector", etc.). But the script always downloads both a gif and gfw file so that the image has a georeference.
If New Zealand distributes its weather data in a way that is similar to NWS, you would just need to modify the script to hit a different website and download different files. If New Zealand doesn't offer a similar distribution method, you would need to write an entirely new script.
Once you modify the current script or create a new one, you would also need to edit the Precipitation Barriers geoprocessing model and swap "Download Radar Data" with your script and change the parameters of the GP tool.
Thanks for the response. Yeah that clears things up now and I can see how that works. New Zealand does have a site with similar maps on it called MetVuw, although it has the land outline and text on the map too, so when it comes to reclassifying it becomes a bit messy. Also, there is no georeference file, so the python script won't be very useful.
However, I'm going to built a model to bring in Metvuw images.