I've added some new HRRR products on weather.utah.edu. In fact, I've completely upgraded our HRRR postprocessing suite over the past couple of weeks, migrating to python and building in a lot of capabilities that should improve reliability and produce far superior graphics than those that have been around now for 20 years. In particular, the code uses Amazon Web Services as the primary access point for the data, with a fall back to the National Weather Service if there's a hiccup. It will auto backfill if products are not produced (assuming our servers are up). It's also parallelized and, for the three product types that are currently being produced, and download and process an 18-hour HRRR forecast in three and a half minutes. The color schemes are also supposed to be more color-blind friendly.
The three products that are available are composite reflectivity and IR, hourly precipitation, and total precipitation.
I wasn't aware that you are responsible for coding some of the weather.utah.edu visuals. Well done! I'm impressed.
ReplyDeleteBravo!
ReplyDeleteVery nice! I really like the grey background. Glad you are aware of the time-height issues...I was very confused yesterday when suddenly I was looking at model runs from May.
ReplyDeleteYes, apologies for that. Everything died after they upgraded our computer system in may as my our old code is apparently too old.
DeleteI have a working prototype to produce the skew-T's and time-heights again. They are really nice, and the code is parallelized and blisteringly fast, but I need to do some more testing and then migrate it to our production suite.