Are random (stochastic) shortcuts better than fixed ones for small-scale processes in climate models?
- Good it would do
- 2 / 5
- Can AI help
- 2 / 5
- How neglected
- 1 / 5
- Can we check it
- 3 / 5
Why we stopped
Rejected on prior art and tractability: stochastic parameterisation is operational at weather centres and actively funded (Oxford's EuroHPC allocation, M2LInES, LEAP), the scout's Lorenz '96 comparison was published in May 2026, and the open climate part (noise versus re-tuning in coupled models) needs cluster-scale runs. A narrow equal-tuning forced-response toy test may be submitted as a new problem.
Why it's stuck
Named as an outstanding open question in applied climate mathematics in 2013 and still unsettled. It sits between pure math and modeling, so neither community owns it.
What we'd try
Run large matched experiments comparing both kinds of scheme against fully resolved simulations, and use AI proof and search tools on the simplified cases.
Think we got it wrong?
Ruled-out problems can come back. Tell us what we missed.
Members can read the full record for this problem, check its sources and comment. Open the member view →