Work for this project has moved here. This repository is kept for archival purposes but is not actively maintained.
This repository contains the code to reproduce the results in the paper: "Tracking emissions in the US electricity system", by Jacques A. de Chalendar, John Taggart and Sally M. Benson. Proceedings of the National Academy of Sciences Dec 2019, 116 (51) 25497-25502; DOI: 10.1073/pnas.1912950116
Note: this setup has only been tested on Linux/OSX. Adjustments may be needed to run on Windows.
- Clone or download this repository to your machine.
- We provide a
setup.pyfile, so you can runpip install ./in the repository root to make sure you have the correct python packages installed. Key dependencies include numpy (>=1.17) and pandas (>=0.25). - This project uses environment variables. They can be set automatically for you using the provided
.envfile (see Usage section), or manually if you prefer to use another location. The environment variables you need are:CODE_PATH: the path to the folder where you downloaded this repository.DATA_PATH: the path where you will save data.FIGURE_PATH: the path where you will create figures.
- Run
source .envfrom the folder this file is in to automatically set the required environment variables.
- We provide the files for the final dataset in the
data/analysisfolder. Figures in the paper (and most of the SI Appendix figures) can be regenerated with these files. Thefiguresfolder contains examples for how to plot and manipulate the dataset. - Alternatively, you can regenerate the dataset, as well as intermediate steps. We provide a Makefile in the
srcfolder to do so. To regenerate the full dataset, you can type the following commands from thesrcfolder:./scrapeAMPD.shto download the AMPD data (depending on the speed of your connection, this can take some time).gunzip ../data/raw/EBA.txt.gzto uncompress the EBA data.make allto regenerate the hourly consumption emissions dataset.
- We provide a Makefile in the
reports/tracking_emissionsfolder to regenerate the figures in the paper and SI Appendix. The following commands can be used from that folder:make main: figures for the main paper.make si_basic: figures for the SI Appendix that do not require building the intermediate data steps.make si: all figures for the SI Appendix.make all: all figures.
- Figures can be also generated one by one by running the notebooks.
- The maps and sankey diagrams can be visualized by opening the html files in the
figuresfolder (underd3_mapandsankeyrespectively).
- Data are stored in the folder pointed by the
DATA_PATHenvironment variable. By default, this is the provideddatafolder. There are three datasets:- AMPD: We download from the EPA ftp server. We provide a script in the
srcfolder to do this (run./scrapeAMPD.shfrom the command line from thesrcfolder). This should go todata/raw/AMPD. - EIA EBA data: available here. The
EBA.txtfile should go todata/raw/. We provide a gzipped version of this file. - eGRID data: available here. The EGRID folder should go to
data/raw/.
- AMPD: We download from the EPA ftp server. We provide a script in the