Preamble

This document explains how to reproduce Figures 1-3, plus Electronic Supplementary Materials Figure 1, from Applegate et al (2015, Climate Dynamics). These figures were produced using output from the three-dimensional ice sheet model SICOPOLIS, development v3.0, as well as a two-dimensional ice sheet model written by Byron Parizek (see references in the paper). We include the model output in this archive; however, we do not provide instructions for running these models here.

Figure 4 was produced by Robert E. Nicholas using climate model output from the 5th Coupled Model Intercomparison Project (CMIP5; http://cmip-pcmdi.llnl.gov/cmip5/). Instructions for reproducing this figure appear in the greenland_amplification_figure_4 folder. The data needed to reproduce this figure must be downloaded from the CMIP5 archive.

We made Figure 5 using the vector drawing program Inkscape v0.91 (https://inkscape.org). We include a .pdf version of the figure in this code package.

Necessary software

The following pieces of software were used to create Figures 1-3, plus Electronic Supplementary Materials Figure 1:

• R v3.1.2 -- statistical computing, figure generation
• RStudio v0.98.1103 -- integrated development environment for R
• Inkscape v0.91 -- figure polishing

R packages

We used the following user-developed software libraries (packages) to generate our results. These packages must be installed using the install.packages() command in R before running any of our R code. Some of these packages depend on other packages; by default, R will install these prerequisite packages automatically.

• RColorBrewer v1.1-2
• signal v0.7-6

Contents of this archive

Directories

The top-level directories in this archive are

• data: contains .csv files with model output from both SICOPOLIS and the two-dimensional ice sheet model
• figures: contains the figure files generated by the R scripts in this archive (see below)
• tables: contains .csv files with quantities derived from the model output (generated by diagnose_derived_quantities.R from the .csv files in the data directory)

.R files

The key .R files are

• diagnose_derived_quantities.R: extracts key quantities from the model output in the data folder, generates tables of these key quantities, and writes them as .csv files in the tables folder
• exp_wsse.R: includes a function, exp.wsse.R, that fits a decaying exponential to a set of points with associated uncertainties in the y-direction
• Fig1_EDFig1_script.R: produces Figure 1 in the main text, plus Figure 1 in the Electronic Supplementary Material; uses exp_wsse.R
• Fig2_script.R, Fig3_script.R: produce Figures 2 and 3 in the main text

.csv files

The .csv files with names that follow the pattern flowline_*.csv represent output from Byron Parizek's two-dimensional ice sheet model. In these files, the first column is time in years. The other columns represent the simulated ice volume in km³ as a function of the input temperature anomaly (0, 3, 4.5, 6, and 12 K, respectively).

The .csv files with names that follow the pattern icevol(t)_dT*_*ka.csv contain output from SICOPOLIS. The number after dT in the filenames indicates the temperature anomaly (0, 1, 2, 3, 4.5, 6, 9, and 12 K), and the number before ka indicates how long the simulations in each file were evaluated (either 100ka, 10⁵ years, or 60ka, 6* 10⁴ years). In these files, the first column is time in years. The other columns represent the simulated ice volume in meters of sea level equivalent for 27 model runs with different input parameter combinations (runs #3, 7, 8, 15, 18, 19, 23, 29, 30, 40, 43, 45, 46, 47, 49, 51, 53, 60, 66, 67, 70, 78, 82, 87, 88, 89, and 92 from Applegate et al., 2012).

The remaining .csv files, which are all stored in the tables directory, give various quantities that were extracted from the SICOPOLIS output files using diagnose_derived_quantities.R. These files are organized with rows indicating temperature anomaly (from top to bottom, 0, 1, 2, 3, 4.5, 6, 9, and 12 K) and columns indicating which model run is being considered (see the last paragraph).

The particular derived quantities stored in each .csv file, and the corresponding units, are as follows:

• deltaV.csv: equilibrium ice volume change (initial ice volume minus final ice volume), in meters of sea level equivalent
• tau.csv: e-folding time of ice volume adjustment, in years
• V_tau.csv: ice volume when the e-folding time is reached, in meters of sea level equivalent

.pdf files

The numbering of the .pdf files is consistent with the figure numbers in the final paper. Note that ESM_Fig_1.pdf is stored on the journal's Web page.

Manual modifications to the figure files

Most of the figures were manually modified using Inkscape before publication. These modifications are not reflected by the .pdf files in this archive (see the published paper for the final figures).

• Figure 1: Fig_1a.pdf and Fig_1b.pdf were combined into a single document. Annotations were added to the top panel of the combined figure to better show the meaning of the e-folding time. In the lower panel, the string "\Delta T_grl =" was added to show that the different colors indicate groups of runs that use the same temperature anomaly.
• Figure 2: The y-axis labels on the top panel were changed to conform with standard ways of displaying information on logarithmic axes. The gray and pink lines and labels were changed to black.
• Figure 3: The curves and points were labeled according to the forcing temperature anomaly.

Running our R code

To reproduce our results, first run diagnose_derived_quantities.R. The other scripts can be run in any order.

To run an R script, first open it in RStudio. Next, click Session > Set Working Directory > To Source Location in the menu at the top of the screen Finally click the Source button in the upper right-hand corner of the Script window in RStudio. Remember to install the packages listed above before running these scripts.