robertoon/misti_inversion: Tephra sedimentation with uncertainty quantification

This release contains the forward numerical model used to simulate tephra sedimentation from umbrella clouds (for details of the model see Constantinescu et al., 2021) and the uncertainty quantification algorithms used to invert the El Misti deposit. The deposit data used here is found in 'field_data.txt' and contains "lat/long/thickness" information for each measured location.

The /misti folder contains the forward model and its dependencies (main.py and ./src folder). The inversion algortihms are found in the same folder (pestpp-glm and pestpp-ies (White et al., 2020)). The './misti_inversion/pest_runs.py' executes the inversion analysis with the parameter ranges defined in pest_runs.py.

The input requires an initial configuration file found in './misti_inversion/misti/src/config.py', the field data found in ./misti_inversion/misti/misti_grainsize.csv' and the definition of the parameters serach domain in './misti_inversion/pest_runs.py'.

To execute the code, run this sequence in 'pest_runs.py':

• 'setup(volcano)'
• 'run_prior_monte_carlo(volcano,num_reals=500,num_workers=10)'
• "run_glm(volcano,noptmax=10,num_reals=0,num_workers=15)'
• 'plot_glm_results(volcano,pmc_dir="{0}_pmc_master".format(volcano))'
• 'plot_glue_results(volcano,glm_dir="misti_glm_master")'