Python code to calculate the distribution of rain and metrics described in Pendergrass and Deser 2017
Starting from precipitation data,
- Calculate the distribution of rain
- Plot the change from one climate state to another
Demo data are included in pdistdemodata.nc
The rain distribution calculations are based on the Matlab shift-plus-increase-modes-demo.
You can read about the methods for calculating the distribution of rain here:
Pendergrass, A.G. and D.L. Hartmann, 2014: Two modes of change of the
distribution of rain. Journal of Climate, 27, 8357-8371.
doi:10.1175/JCLI-D-14-00182.1.
The response to warming is described in:
Pendergrass, A.G. and D.L. Hartmann, 2014: Changes in the distribution
of rain frequency and intensity in response to global warming.
Journal of Climate, 27, 8372-8383. doi:10.1175/JCLI-D-14-00183.1.
The rain amount peak, rain frequency peak, rain amount width, and rain frequency width metrics are described in a forthcoming paper (currently being revised) for the Journal of Climate by Pendergrass and Deser.
Please cite one or all of these papers.
A bit of sample data from a member of the CESM1 large ensemble is included to get you going. (https://www2.cesm.ucar.edu/models/experiments/LENS)
You can also use your own gridded precipiation dataset, of course. For example, you can use daily rainfall data from a CMIP5 simulation (GFDL-ESM2G is shown below, for the 1pctCO2 scenario) which you might be able to download, for example from PCMDI (https://pcmdi.llnl.gov/projects/cmip5/)
Or you can use data from GPCP 1dd https://climatedataguide.ucar.edu/climate-data/gpcp-daily-global-precipitation-climatology-project or TRMM 3B42 https://climatedataguide.ucar.edu/climate-data/trmm-tropical-rainfall-measuring-mission.
Get in touch if you have questions, or if you're interested in collaborating.
20 January 2017, Angeline Pendergrass, NCAR, Boulder CO. apgrass@ucar.edu