MATLAB Code used in: Loss of Earth System Resilience during Early Eocene transient Global Warming Events

This package has MATLAB codes to run analyses performed in Setty et al. Loss of Earth System Resilience during Early Eocene Global Warming Events. It includes 3 .zip files which contains the ccm toolbox,  generic early warning signals toolbox, and the analyses performed in the manuscript. To run null models for DCCM and DRCC, ebisuzaki.m and sievebootstrap.m are required.

The data used for the analysis in the paper are available in permanent repository https://doi.pangaea.de/10.1594/PANGAEA.897741 and https://doi.pangaea.de/10.1594/PANGAEA.893894.

The RCC MATLAB code applied here in a sliding window and the code is published on ResearchGate (https://www.researchgate.net/publication/355033193_A_matlab_programing_code_for_the_Reservoir_Computing_Causality) and the paper (Huang Y, Fu Z, Franzke CLE. Detecting causality from time series in a machine learning framework. Chaos. 2020;30(6)). This code uses the Echo State Network Toolbox which is publicly available on https://www.ai.rug.nl/minds/research/esnresearch/.

General description of the codes in Manuscript_analyses:

data_curation.m  includes loading and selecting the datasets to perform DIORs trend analysis and DCCM and DRCC analysis
dior_analysis.m  includes using generic_ews toolbox to perform sliding window DIORs analysis, sensitivity analyses and creating Figure 2
dccm_to_run.m is the DCCM code along with the code to create Figure 3. 
rcc_to_run.m is the DRCC code 
dynamic_causality.m includes code to create Fig S12 and Fig. S13
convergence_test.m is the code to perfom the convergence test
eccm.m is the ECCM code  and to create figures
data_transform_SA.m tests the sensitivity of DCCM analysis to different data transformation 

This package also includes results to create Figure 3, Fig. S7, S12, S13.