Why Austria belongs in the first expansion batch
The Tirol valleys are the country where the case for a regional model writes itself. Here is why Austria sits in the first European batch.
Austria is the country in the European batch where the case for running our own regional model writes itself. This is a quick note on why it belongs in the first batch.
The Tirol valleys do not share a morning inversion - launch on Bramberg can be working an hour before launch on Wildkogel forty kilometres away. A global model averaging across both is wrong about both. Föhn is a synoptic-to-mesoscale interaction that needs real terrain to compute. Convergence lines along the main Alpine ridge are exactly the kind of feature a 4 km grid resolves and a coarser model parameterises into mush. Most of what makes Austrian flying interesting only exists at this scale.
The compute cost is the largest in the European batch so far - the Austria rectangle is longer than Switzerland and benchmarks at around 23 minutes per cycle. That is the country that started to make the multi-country scheduler earn its keep. It still fits comfortably inside four cycles a day with the rest of the batch on the schedule, which is what the staggered timer design was for.
Austria also makes the product more obviously Alpine. Switzerland proves dense mountain coverage, Slovenia proves a small travelling-pilot country, and Austria proves the scheduler can carry a longer Alpine domain without changing the API shape.