RASP - Improving our Soaring Forecast

Output of a typical RASP run.
Source: Valley Soaring.net

"Do you know what RASP is?"  It was a question from one of our best sailplane racing pilots.

I told him I didn't really know what it was.  It turns out to be a soaring prediction model called Regional Atmospheric Soaring Prediction, or RASP.   The RASP model turns out what are called blipmaps which are maps of soaring conditions that indicate all kinds of parameters of interest to soaring pilots.

These maps are already available online at www.drjack.com, but the grid areas are large and the terrain isn't considered as precisely as it could be.  Dr. Jack Glendening is an atmospheric scientist and the author of RASP.  He allows gliding clubs to use his code at no charge, but also with community-only support.

I took a look at RASP, which is a collection of Unix programs that are controlled from an overall RASP script.  It is a terminal only program - there is no GUI, and all output during the runs is written console-style to logs that you pore through to determine the source of your errors.  Parameters are in different directories and not always obvious.  The documentation for the program is fairly robust, but suffice it to say that troubleshooting a complex software package that you have never seen before is quite a challenge.

So far, I've managed to get RASP right up to the edge where it actually run the atmospheric forecasting program.  In effect, the program so far downloads the results of model runs from the National Weather Service, then preps them for a local run of WRF, the Weather Research and Forecasting model.  That's where the hard number crunching happens, then the results are output in a format you can put them on a website with graphical charts showing the results.

I've come quite a way to make just a little progress and I'm currently stumped at one error.  The forum for RASP is understandably small and Dr. Jack doesn't actively support RASP, so if I can't figure it out, I'm sunk.  I've got one or two ideas right now, but right now I'll give myself about a 70% chance of success.