1/n Earth system models are not 'just theory'.

Models rest on observation and experiments that reach back over centuries. (Yay Tyndall!)

Models use to assess solar geoengineering have errors that can be reduced by experiments.... https://twitter.com/CommsEarth/status/1338808766075199489

2/n @ZhenDai30's paper concludes with three examples of other straightforward examples of experiments that could reduce important uncertainties about predictions for solar geoengineering aka solar climate intervention.

Concluding section.

Final para below

3/n Experiments are hard. They take a long time. They are messy. These results show that our paper on Calcium Carbonate and ozone exaggerated the ozone impacts.

Our experiment shows that our model was wrong.

4/n Some argue that physical science work on solar geoengineering climate intervention should concentrate on applying existing models.

Exclusive reliance on models promotes overconfidence.

Our case for empirical research prepared for the NAS:

https://keith.seas.harvard.edu/files/tkg/files/keith_keutsch_floerchinger_emperical_srm_for_nas.pdf