This archive contains raw CESM output data for:

1) All 3-day simulations (initialized hindcasts) for the Morris sensitivity analysis
2) The baseline and targeted perturbation 20-year simulations
3) The two single-perturbation runs for 10 years (only single parameters perturbed)

The raw CESM output data is on the native spectral element (SE) grid. Only data for t=72 hours is available in this archive. The 3-day simulations contain instantaneous output (h0), daily-average output (h1), and instantaneous flux budget output (h2). The long-term simulations have monthly-average data (h0) and daily instantaneous data (h1, for the single-perturbation runs). 

Each CESM configuration has its own subdirectory. The 4000+ 3-day simulations are labeled ${initialization date}_mm5.${ensemble number}. The long-term simulations are labeled as follows:

10yr_baseline: 20-year baseline simulation ("10-year" isn't a typo, apologies for the confusing label)
10yr_test2: 20-year targeted perturbation simulation ("10-year" isn't a typo, apologies for the confusing label)
10yr_test5: 10-year simulation with only C_invrs_tau_shear perturbed
10yr_test6: 10-year simulation with only C8 perturbed

This archive also contains analysis data that is used to make all plots in the main text and supplement. There's also a directory with all necessary code to re-create these results. We note most simulations were performed on NCAR's Cheyenne supercomputer, which is no longer active. The code should run with few modifications on the Derecho supercomputer with nearly identical results, but this hasn't been verified.

-------------------------------------
NOTE: The label "mm5" does not refer to the regional model of the same name. This naming convention was adopted by the authors for organizational purposes. 
-------------------------------------