Global lake synchrony amplifies planetary boundary risks

Zenodo Description

Title: Dataset and Code for: Global lake synchrony amplifies planetary-scale risks

Description: This repository archives the processed datasets and source codes associated with the research article: " Global lake synchrony amplifies planetary-scale risks ", submitted to Science.

This study integrates long-term remote sensing data to investigate the spatiotemporal synchrony, network modularity, and stability mechanisms of global lake ecosystems.

1. Raw Datasets

The raw input data includes long-term monthly time series, primarily derived from satellite remote sensing products.

• Global_Lake_Water_Color_Time_Series.csv: Monthly Forel-Ule Index (FUI) time series.
• Global_Lake_Chlorophyll_a_Time_Series.csv: Monthly Chlorophyll-a concentration time series.

2. Source Data for Figures (Processed Data)

This section contains the processed numerical data generated by the analysis scripts, which were used to create the figures in the manuscript.

Figure 1.rar (Related to Figure 1):

• Content: Geospatial source data used to generate the study area map. Contains the GIS project file and associated vector/raster layers.
• Format: Compressed archive (.rar).
• Software Requirement: Requires ESRI ArcGIS (ArcMap or ArcGIS Pro) to open and view the map project.

Figures_data.csv:

• Content: Consolidated source data for the majority of the figures in the main text and Supplementary Information.
• Exclusions: Source data for Figures S5, S6, S8, and S9 are not included in this file. These figures visualize dynamic results and are generated directly by the provided Python scripts.

3. Code Availability

The repository includes a comprehensive Python workflow for ecological network construction, synchrony analysis, and statistical validation.

A. Spatial Network & Structure

• Lake_Network_Modularity_Analysis.py:

Constructs lake similarity networks based on time-series correlation; performs community detection using the Louvain algorithm; and conducts sensitivity analysis on modularity resolution.

B. Temporal Synchrony & Dynamics

• Lake_Synchrony_Analysis.py:

Calculates the population spatial synchrony index, incorporating data preprocessing steps such as STL trend decomposition and frequency filtering.

• Lake_ Correlation_Analysis.py:

Investigates the temporal evolution of mean pairwise Pearson correlations. The workflow integrates robust data preprocessing (interpolation, frequency filtering, and STL detrending) and evaluates the significance of long-term correlation trends using linear regression with 95% confidence intervals derived from bootstrapping.

Note: Due to the edge effects inherent in the sliding window and detrending algorithms, the first 12–24 data points of the output series are considered initialization artifacts and should be manually trimmed for robust analysis.

C. Robustness & Significance Testing

• Lake_Synchrony_Sensitivity_Analysis.py:

Validates the robustness of synchrony results by testing across multiple sliding window sizes (e.g., 2, 5, 8, 10, 15 years) to ensure conclusions are not artifactual.

• Null_Model_Generator.py:

Generates 1,000 randomized surrogate datasets by independently shuffling the original time series. This constructs a "null model" baseline by destroying temporal structure while preserving data distribution.

• Lake_Synchrony_Significance_Test.py:

Performs statistical significance testing (Mann-Kendall) by comparing observed synchrony trends against the null distribution derived from the randomized datasets (calculating empirical p-values).

4. System Requirements

• Language: Python 3.8
• Dependencies: numpy, pandas, scipy, statsmodels, networkx, matplotlib, pymannkendall, scikit-learn, python-louvain, tqdm.