Simulating Earth with 1.25x solar flux

I have tried changing the solar flux to be 1.25x that of Earth by doing: ```julia spectral_grid = SpectralGrid() flux = 1.25 * 1365 p = Earth(spectral_grid, solar_constant = flux) model = PrimitiveWetModel(spectral_grid; planet=p) ``` After around 2.5 years the model blows up. When I run the same code but for Earth: ```julia spectral_grid = SpectralGrid() flux = 1365 p = Earth(spectral_grid, solar_constant = flux) model = PrimitiveWetModel(spectral_grid; planet=p) ``` No blow-up occurs. Some interesting things I noticed: (thanks @nviebig for the code to help me do this btw!) 1. Radiation New planet radiation budget <img width="2400" height="1600" alt="Image" src="https://github.com/user-attachments/assets/a6ac8e74-84d0-47f7-adc0-4a59bcc2da27" /> Earth radiation budget <img width="2400" height="1600" alt="Image" src="https://github.com/user-attachments/assets/ae5452ca-83fa-48b0-a48e-272bb51beae4" /> New planet: - The outgoing long wave radiation steadily increases - Net top-of-atmosphere radiation is huge - Absorbed shortwave radiation is also huge - Surface downwards shortwave radiation is decreasing Earth: - Earth seems stable but still way too much shortwave absorption - I think too much shortwave radiation is being absorbed? Hence why TOA net radiation is so high (not much is being taken out!) 2. Temperature New planet: <img width="2200" height="1000" alt="Image" src="https://github.com/user-attachments/assets/a8a55fdc-abc6-4c3a-b120-db721b31050e" /> <img width="800" height="900" alt="Image" src="https://github.com/user-attachments/assets/7771c87c-ab33-49af-a6e3-9dceaa096edb" /> Earth: <img width="2200" height="1000" alt="Image" src="https://github.com/user-attachments/assets/88316e84-b249-42e6-aaec-e957e52fdae7" /> <img width="800" height="900" alt="Image" src="https://github.com/user-attachments/assets/1ac79668-98d1-4c5e-a39c-db014514f8de" /> - Earth's temperature profile seems to stabilise in the top layers with a nice inversion in the uppermost layer - The new planets upper layer temperatures are huge - There are 2 inversions in the new planet's temperature profile - My guess would be because it's in the upper layers, ozone is absorbing too much radiation. For my new planet case I'm guessing the higher intensity of incoming solar radiation is causing this problem. I'm not sure why the uppermost layer then inverts again from the layer below though. 3. Dynamics New planet: <img width="2400" height="1600" alt="Image" src="https://github.com/user-attachments/assets/9a8c9439-b51e-492c-996a-e8b8e6002dc9" /> Earth: <img width="2400" height="1600" alt="Image" src="https://github.com/user-attachments/assets/93cfdc05-f9dd-4d92-8404-9874d7302ad8" /> - Vorticity is approx 10^-5 which is OK and divergence is approx a magnitude larger as required for both planets - Vorticity for the new planet is gradually increasing, showing an early onset of an instability - I'm wondering if the increase in temperature issues are causing this? 4. Humidity New planet: <img width="2200" height="1000" alt="Image" src="https://github.com/user-attachments/assets/2620c001-fc90-4f3c-9490-3fbe2070e3a0" /> Earth: <img width="2200" height="1000" alt="Image" src="https://github.com/user-attachments/assets/97fcc32a-1ee5-43c9-9e19-7b44411c4594" /> - All layers in my new planet increase in humidity - I'm guessing the input radiation is causing this as the temperature in each layer will be larger 5. Precipitation I found when calculating the fraction of convective rain/total rain via: ```julia # precipitation: m/s → mm/day scale = 86_400f0 * 1000f0 cb.precip_total[i] = global_mean(diagn.physics.rain_rate) * scale cb.precip_conv[i] = global_mean(diagn.physics.rain_rate_convection) * scale cb.precip_ls[i] = global_mean(diagn.physics.rain_rate_large_scale)* scale ``` I get ```rain_rate_convection``` > ```rain_rate``` This is true for both planets New planet: <img width="2200" height="1600" alt="Image" src="https://github.com/user-attachments/assets/10891f5c-fb74-4634-bae8-afb19f8fc66c" /> Earth: <img width="2200" height="1600" alt="Image" src="https://github.com/user-attachments/assets/bef29564-869a-49e5-bac6-131614b5d3a1" /> Here we can see the fraction of convective rain is consistently over 100% for both planets which shouldn't be possible Overall thoughts: - I have changed the advection scheme to ```WENOVerticalAdvection``` which seemed to work but I'm ideally looking for something less computationally expensive - Changing convection to dry instead of moist stabilised the scheme for a bit longer but blowups did still occur (I haven't got diagnostics for this yet but working on it) Any other thoughts would be very welcome, thank you :))