· Summary

o Addresses the question: Do calderas preserve the shape of underlying magma chambers?

o Models a mobilized and locked magma bodies with a shifting yield surface.

o Performed a Monty Carlo simulation with several free parameters.

o Found that calderas can form from bot complete and incomplete mobilizations.

o Calculate the stress at the magma chamber boundary

· Did they make a good case?

o Why don’t they worry about mixing?

§ Discussion of the timescales involved and that mixing happens on much larger timescales

o Question whether the Fish Canyon Tuff was a good example site for full caldera collapse

§ Perhaps the Fish Canyon caldera fits into one of the bands seen in Fig 2

§ Fish Canyon is very well mixed. Paper says the mixing happened prior to the event and is not a concern for the models

· Do you think the choice of scope is appropriate? / Why did this work end up in a high profile publication?

o The yield stress component makes it unique and opposed to models where the entire chamber is mobilized.

o Challenges current thoughts using a dynamic model

· Discussion on why people think the whole magma chamber is evacuated in caldera collapse.

o Discussion of what our current volcano model is.

· Why does the magma chamber have the shape it does in the model?

o As you increase in volume it I easier to add mass laterally

· Why doesn’t caldera collapse in a partial ejection system remobilize the magma?

· Discussion of Fig 2.

o Model can say a lot but caution should be used when interpreting it. Agree they did a good job.

o They don’t get above 50% of the erupted volume

· How can we account for sequential eruptions giving different chemical compositions?

· Discussion of whether the need for strong country rock reflects reality of know sites

· Why doesn’t varying conduit change things?

o Choked flow condition assumes flow at the sound speed

o Does this assumption make sense?

o Model assumes isothermal conditions but model parameters include temperature rates.

· How do you decide how many runs to do in a Monty Carlo simulation?

· What is Kolmogorov-Smirnov test?

· Feedback: How do people feel about the format/style of the paper