Modeling “Bloopers”

This page is provided for your entertainment 🙂 .

Sometimes in models, things go horribly, terribly wrong…

Parallelization in SLIM

Early in the development process…

How many South Americas can YOU spot?

Split the simulation up over multiple, parallel processors? Check! Put it back together again? Nope, not check.

The best part of this was that it didn’t crash.

Units? What are those?

When developing SLIM, I did all the energy-conservation checks on evaporation (and the surface energy budget in general) on the variables in energy units (W/m2). Then I converted evaporation to water units to pass up to the atmosphere, in the processes including an extra factor of 1000 just… for fun? (By accident). Turns out CAM will throw an error if the clouds reach the top of the model!!!! Good thing…

600,000,000 ppm CO2

This looks fine, right? Until you look at the colour bar… yes, the surface temperature has reached 380 K.

Welcome to the hot house. Pro tip: if you’re changing CO2 in CESM, check your units! Setting CO2 in the CAM namelist? Use mixing ratio (parts per… part). Setting CO2 in the CLM namelist? Use ppm. Oopsies. CAM5 will run for 12 months at 600,000,000 ppm. CAM6 will run for 6 years!