Skip to content

Latest commit

 

History

History
53 lines (35 loc) · 2.22 KB

File metadata and controls

53 lines (35 loc) · 2.22 KB

Robust Adaptation to Multi-Scale Climate Variability

DOI download the exact version of this code used to create our manuscript

Code style: black

Welcome to the code repository for the paper "Robust Adaptation to Multi-Scale Climate Variability" by James Doss-Gollin, David Farnham, Scott Steinschneider, and Upmanu Lall, under review in Earth's Future. Running the code in the src directory will enable you to generate all the figures in our paper, including supplemental figures. You can also view original versions of several presentations of (early versions of) this work. Please see the published (open-access) version of this paper for final text, figures, and references.

Code and Installation

All the code except that which generates the wavelet spectrum of the NINO3 annual series is written in python and lives in the src directory. You can install all the required packages in conda:

conda env create -f environment.yml
conda activate robust-adaptation-cyclical-risk

Running

There are two steps to running the codes.

The first step is to run the computational experiments. The relevant files are called slurm-LFV-Only.sh, slurm-LFV-Secular.sh, and slurm-Secular-Only.sh. These files are currently written to be run on Columbia's Habanero cluster, but you can easily run them elsewhere. Simply replace the comments at the top of each file (these are for a slurm scheduler) and replace

module load anaconda
source activate robust-adaptation-cyclical-risk

with

conda activate robust-adaptation-cyclical-risk

or whatever your system requires. Running this code will take a long time and will generate a lot of data (order 10-100GB), so be prepared! Once the experiments have run, you can use jupyter notebooks to visualize results. These also live in the src directory and are numbered -- just run them in order.

Issues

For any problems running the code, please open an issue in the issues tab. For general questions, please contact James Doss-Gollin