Data & code repository for Forced changes in the Pacific Walker circulation over the past millennium This repository contains scripts and auxiliary data necessary for recreating the figures in the paper Forced changes in the Pacific Walker circulation over the past millennium (https://doi.org/10.1038/s41586-023-06447-0) [published in Nature October 2023; available online August 2023]. The code is tested with R version 4.1.2. Note that the only difference between this and the previous version is a slight update to the reconstruction filenames Repository Structure `Falster_PWCreconstructions`: zipped folder contains our ensemble reconstruction of ΔSLP (1200-2000 CE) `Falster2023_PWC_reconstruction.csv`: a csv with one column per ensemble member. Column headers state the reconstruction method, and gridded SLP product used to train the reconstruction (200 ensemble members per unique combination of reconstruction method and training index). `Falster2023_PWC_reconstruction.xlsx`: Excel workbook, with one tab per unique combination of reconstruction method and training index (200 ensemble members each). `Falster2023_PWC_reconstruction_full_ensemble_median_and_95pct_range.csv`: a csv with the median and 95% range (2.5th and 97.5th percentiles) of the full 4800-member ΔSLP reconstruction ensemble. Includes raw summary values, and summary values with a five-year running mean. For robust analysis of ΔSLP, we strongly recommend working with the full ensemble rather than these summary values. If you need tips on working with ensemble data, please get in touch with Georgy Falster. `Falster_PWCreconstruction_code`: zipped folder contains scripts and auxiliary data necessary to generate figures in the main text `scripts`: contains R scripts necessary to generate figures in the main text `Figure1`: folder containing scripts to perform analysis and generate Fig. 1 in the main text `Figure2`: folder containing scripts to perform analysis and generate Fig. 2 in the main text `Figure3`: folder containing scripts to perform analysis and generate Fig. 3 in the main text `Figure4`: folder containing scripts to perform analysis and generate Fig. 4 in the main text `Figure5`: folder containing scripts to perform analysis and generate Fig. 5 in the main text `Figure6`: folder containing scripts to perform analysis and generate Fig. 6 in the main text `data`: contains all auxiliary data necessary for the analysis and visualisation, including: reconstructions in .rds format shapefiles for mapping processed ΔSLP and Nino 3.4 RSST timeseries from the CESM1 LME and PMIP3/4 models How to cite this repository If using code or data from this repository, please cite the original publication, available from https://www.nature.com/articles/s41586-023-06447-0. This repository can be cited with https://doi.org/10.5281/zenodo.7742760.

Production of this code and dataset (and the associated paper) was supported by US National Science Foundation (NSF) grants AGS-1805141, AGS-1805143, AGS-2041281, and OCE-2202794, as well as the Australian Research Council Centre of Excellence for Climate Extremes. The Iso2k database used in production of this dataset is a contribution to Phases 3 and 4 of the PAGES 2k Network. PAGES receives support from the Swiss Academy of Sciences, the US NSF and the Chinese Academy of Sciences.