Press release about our ongoing forecast demonstration with a global 3.75-km MPAS:
news.ucar.edu/133040/scien...

Four days before Typhoon Ragasa’s rainbands hit Taiwan, the 3.75-km MPAS predicted >36" of rain in the mountains—while the 15-km parallel run only showed up to 24".

By better resolving Taiwan’s terrain and upslope rainfall, the high-res run shows the value of higher resolution in complex terrain.

Four days before Typhoon Ragasa’s rainbands hit Taiwan, the 3.75-km MPAS predicted >36" of rain in the mountains—while the 15-km parallel run only showed up to 24".

By better resolving Taiwan’s terrain and upslope rainfall, the high-res run shows the value of higher resolution in complex terrain.

While our experimental 3.75-km MPAS forecast did not move the storm as quickly as in reality—placing it farther east and farther from land—it nevertheless captured a super typhoon before the system was even named #Ragasa
👉 project.mmm.ucar.edu/real-time-fo...

18-Sep 00Z MPAS forecast has a major typhoon aiming at the Luzon Strait in three days. This system hasn't even formed yet, so we'll see.
👉 project.mmm.ucar.edu/real-time-fo... Variable: Infrared Brightness Temp, Domain: Invest 90 (WP90)

Today's tidbit from @ncar-ucar.bsky.social's experimental 3.75-km MPAS forecasts: An active train of African easterly waves.
- a squall line over western Africa
- a tropical storm near 30°W
- another less well defined tropical system near 45°W
👉 project.mmm.ucar.edu/real-time-fo...

While a Mesoscale Convective System did form this evening over the Colorado plains (left), the MPAS forecast had it farther north than observed (right)

Today's #MPAS forecast has two cyclones in the eastern Pacific that appear to be locked in a Fujiwara-like fashion.
👉 project.mmm.ucar.edu/real-time-fo...
variable: 10-m Winds, domain: EPAC (zoom), forecast hour 120

One advantage of high-resolution global models is their ability to resolve topography and its influence on weather. In this 5-day accumulated rainfall forecast for Puerto Rico, the model predicts little to no rain along the southern coast but more than 200 mm (8 inches) in the northwestern interior.

MPAS brings Hurricane Kiko close to Hawaii (as a 45-kt tropical storm) in ~4 days creating some interesting topographical flow features
👉 project.mmm.ucar.edu/real-time-fo...
Variable: 10-m Winds, Domain: Hawaii

🌍 The MMM lab at @ncar-ucar.bsky.social is running experimental realtime forecasts with a global 3.75-km MPAS model.

Goal: Highlight the power of global km-scale models for weather prediction from local to planetary scales, incl. tropical cyclones.

👉 project.mmm.ucar.edu/real-time-fo...

Tired of guessing how many core-hours your MPAS run will devour? My new script (github.com/falkojudt/mp...) helps you to get the numbers.

Trade wind cumulus at work, transporting moisture from the boundary layer into the free troposphere

The downslope flow drove temperatures into the mid to upper 60s (~18C) on the eastern slopes with little humidity. There’s still some snow cover over Colorado the high plains, which creates some local cold patches.

Down near the surface, there was a lee trough east of the front range, which helped to drive downslope flow.

In the upper levels, there’s a ridge over Colorado, westerly flow, and an upstream a trough over the Pacific. This set up brings in some upper level moisture which often produced wave clouds, which we had today (though the clouds broke over Boulder).

A beautiful mild late autumn day by Boulder creek.
So, what’s going on meteorologically? #cowx

Great example of a “cyclone train”, where each storm takes a similar path to the previous one.

Big question is: are these “trains” predictable on S2S time scales (weeks in advance)?

Flying is such a privilege. I can spend hours 👀 all the landmarks and the weather from above…
(1) Golden, CO and its table mountains
(2) Berthoud Pass
(3) Rocky🏔️ after the first ❄️, barren South Park in the background (Copper ⛷️ peaks out, too)
(4) Fog moving into dry hills near San Francisco