Overview

RESCUES (Resilience Enhancement in Supply Chains Under Environmental Shocks) is a supply chain network-based framework to evaluate both the direct and ripple indirect economic risks posed by individual and compound hydroclimate shocks under future climate change based on the Shared Socioeconomic Pathways (SSP) scenarios.

Usage

The scripts should be executed in the following order:

1. Extract_region_level.py

This script processes high-resolution precipitation data from CMIP6 climate models and aggregates it to the national level. It calculates area-weighted monthly precipitation for each country based on a country mask file and grid cell areas. The script supports multiple models and SSP scenarios (e.g., SSP1-2.6, SSP5-8.5), and outputs a CSV file of precipitation time series by country.

Key features and steps:

• Computes the surface area of each grid cell.
• Calculates area weights for each country based on its spatial extent.
• Aggregates gridded precipitation data to country level using the area weights.
• Supports batch processing of multiple climate models and scenarios.
• Outputs monthly precipitation (mm/month) for each country from 2015–2100.

2. ag_losses_extreme_spi.py

This script estimates agricultural productivity losses caused by hydroclimatic extremes—specifically, compound droughts and pluvials—using area-weighted standardized precipitation indices (SPI) derived from CMIP6 precipitation projections. It combines climate model outputs with empirical coefficients from regression results (Kotz, 2022) to quantify the marginal economic effects of extreme weather on agriculture at the country level.

Key features and steps:

• Converts gridded monthly precipitation data into country-level SPI using WASP (Weighted Anomaly Standardized Precipitation).
• Categorizes SPI values into nine types of extremes (extreme/severe/moderate wet, dry, and compound).
• Applies estimated marginal effects to compute agricultural loss for each country and year (2015–2054).
• Outputs annual country-level agricultural losses for each climate model and SSP scenario.
• Aggregates multi-model means for each scenario and saves results as CSV files.

Output:

• Country-year SPI and WASP tables.
• Agricultural loss estimates by shock type (e.g., w_losses_pExtreme, w_losses_comSevere, etc.) under each SSP.
• Multi-model mean loss estimates for comparison.

3. simuTechProExtreme.py

This script simulates the cascading economic impacts of agricultural productivity shocks—caused by hydroclimatic extremes—on global production.

Key features and steps:

• Integrates scenario-specific agricultural shocks (from climate models and SPI-based loss estimates) into a global multi-regional input–output (MRIO) model.
• Simulates annual shock propagation over a 35-year horizon across 189 countries and 26 sectors.
• Incorporates adaptive behaviors, such as: technological adjustment in production coefficients, stock-based buffering and overproduction, inventory-driven reordering, demand redistribution through evolving final and intermediate input shares.
• Iteratively computes constrained production (IOX_t), value-added (IOV_t), order matrix (OrderMat), and inventories at each time step.

Input requirements:

• Processed agricultural loss data (e.g., w_losses_pExtreme_ACCESS-CM2_ssp585_ag.csv)
• Baseline IO data from Eora26 (intermediate, final, and value-added matrices)
• Auxiliary functions from FUNC.py and data loading utilities in process_data.py

Output files include:

• *_IOX.csv: simulated annual total outputs under shocks.
• *_DirectDamage.csv: direct damage estimates by sector and region.
• *_order.npy: dynamic bilateral inter- and intra-country order matrices for agriculture and all sectors.

4. simuTechProExpResExtreme.py

This script simulates the compounded global economic impacts of agricultural productivity shocks (due to extreme precipitation events) in conjunction with export restrictions on agricultural products. It extends the dynamic economic network framework by incorporating temporary, country-specific export barriers.

Key features and steps:

• Implements export restrictions (e.g., 50% reduction in agricultural exports) for selected countries (e.g., China, USA, India, Indonesia) during specified time windows (two years or five years).
• Integrates country-level precipitation-driven agricultural loss estimates (from SPI-based model outputs).
• Runs multi-year (2016–2050) simulations for all combinations of:
• • 12 climate models
• • 2 SSP scenarios (SSP1-2.6, SSP5-8.5)
• • 9 extreme weather categories (pluvial/drought/compound, in three severities)
• • 5 * 2 export restriction scenarios
• • Captures production adjustments, trade reallocations, and inter-sectoral feedbacks under joint weather-trade shocks.

Output files include:

• *_IOX.csv: Simulated annual total outputs under shocks.
• *_order.npy: Dynamic bilateral inter- and intra-country order matrices for agriculture and all sectors.

5. UA_simuTechProExtreme.py

This script performs uncertainty analysis of global economic output under climate-induced agricultural shocks. It quantifies the impacts of extreme precipitation events on production systems and assesses the resilience of global supply chains by varying key adaptive parameters.

Key features and steps:

• Implements a sensitivity analysis based on six uncertain parameters:
  1. Production technology adjustment: maximum technical progress rate ($\alpha_{js}^{\max}$), adjustment time scale ($\tau_\alpha^\downarrow$)
  2. Demand adaptation: demand elasticity ($\eta_{ir}$), adaptation time scale ($\tau_o$)
  3. Overproduction capacity: maximum overproduction rate ($\beta_{ir}^{\max}$)
  4. Inventory utilization: inventory duration ($n_{i\rightarrow js}$)
• Samples parameter combinations using Latin Hypercube Sampling (LHS) to generate multiple simulation trajectories.
• Simulates 35-year dynamic adjustment processes under multiple hydroclimatic extreme scenarios (e.g., pExtreme, nExtreme, comSevere)

Output files include:

• "./Results/UA_TechPro/IOXExtreme/{scenario}_{simu}.csv": Simulated annual total outputs under each uncertainty scenario.

6. Codes for visualization

• Plot_IOV_Extreme.ipynb: generates Fig. 2 and Fig. S1.
• Plot_per_event.ipynb, Plot_per_event_region.ipynb: Generates Fig. 3 and Fig. S2.
• Plot_dir_vs_ind_extreme.ipynb, Plot_dir_vs_ind_severe.ipynb, Plot_dir_vs_ind_moderate.ipynb, Plot_ineq_gini.ipynb: generate Fig. 4a,c,e; Figs. S3-S7a,c,e.
• Plot_ineq_ua_dir_vs_ind_extreme.ipynb, Plot_ineq_ua_dir_vs_ind_severe.ipynb, Plot_ineq_ua_dir_vs_ind_moderate.ipynb: generate Fig. 4b,d,f; Figs. S3-S7b,d,f.
• Sec_heatmap.R: generates Fig. S8.
• Plot_ExpRes_Extreme_tot.ipynb, Plot_ExpRes_Severe_tot.ipynb, Plot_ExpRes_Moderate_tot.ipynb: generate Fig. 5a,b,c; Figs. S9-S13a,b,c.
• Plot_ExpRes_Extreme_income.ipynb, Plot_ExpRes_Severe_income.ipynb, Plot_ExpRes_Moderate_income.ipynb: generate Fig. 5d,e,f; Figs. S9-S13d,e,f.
• Plot_ExpRes_Combined_5_sectoral.ipynb: generates Figs. S14 and S15.
• Plot_ag_trade_Combined_5.ipynb: generates Figs. S16, S18, S20, S22, and S24.
• Plot_ExpRes_AG_Combined_5.ipynb: generates Fig. 6a-f.
• Plot_ExpRes_AG_USA_5.ipynb: generates Fig. S17a-f.
• Plot_ExpRes_AG_CHN_5.ipynb: generates Fig. S19a-f.
• Plot_ExpRes_AG_IDN_5.ipynb: generates Fig. S21a-f.
• Plot_ExpRes_AG_IND_5.ipynb: generates Fig. S23a-f.
• Plot_ag_trade_ExpRes_dev.R: generates Fig. 6g,h,i; Figs. S17, S19, S21, S23g,h,i.
• Plot_UA.ipynb: generates Fig. S25.