The entrainment analysis toolkit is a package used to post-process output data generated by SAM (System for Atmospheric Modelling) with entrainment calculation implemented by Dawe and Austin (2011a). This package combines (parallelized and automated) SAM output conversion (bin3D2nc
), cloudtracker and several different post-process scripts.
The raw .bin3D
data from SAM is converted, and sorted to be used by cloudtracker. Then for each cloud parcel, the package creates a netCDF
file containing all the relevant output data from SAM. Now with the individual cloud data collected by cloudtracker
, it is possible to apply reanalysis scripts (or analysis modules) to produce a more detailed picture of each cloud parcel -- these modules will modify output netCDF
files and add additional statistical variables needed for the entrainment analysis.
ent_analysis
is now in production.
(12/10/2013) The development of the analysis modules is more or less done (except maybe a number of bug fixes and the introduction of a better parallelization scheme for
id_profiles
andsurface_profiles
, which is a little more difficult due to the way they are written).Unless otherwise requested, I will most likely not include the utility scripts in the package.
(01/17/2014) No more parallelization will be introduced, as the performance of this package does not significantly improve the speed of the analysis (except the cloud tracking algorithm, but that would entirely be another research). But I will keep adding small improvements to this analysis package.
- Automatically read dimensions from input files
- Implement time module to measure execution time
- Integrated data output folder
- Ensure no data contamination by re-run
- Parallelize cloudtracker module
- Implement deeper parallelization for
id_profiles
To run the entrainment analysis toolkit, the following Python modules must be installed (as needed for the cloudtracker module):
- numpy
- networkx
- netcdf4-python or pupynere
Download ent_analysis package to SAM directory, or where the model output will be stored for better performance (recommended if the storage is limited, as the entrainment analysis package will also store the processed data). Ensure that the configuration file config.cfg
is properly modified according to the system configuration.
To run entrainment analysis toolkit, simple run:./run_analysis.py
Or, use the MOAB script msub run.pbs
The post-processed output files are in /time_profiles/cdf
and in /id_profiles/cdf
.