Geopolitical-adaptive shipping contracts: simulation model and results (v2b)
Year: 2026
Type: software
Abstract: Companion deposit for the manuscript "Geopolitical-adaptive shipping contracts: a signal-indexed mechanism for continuity under compound maritime disruption" (Transportation Research Part E, special issue on resilient and intelligent maritime transportation systems under geopolitical uncertainty). Version v2b, 5 October 2026. The deposit contains the complete simulation model and every result reported in the paper. gasc_model.py implements the geopolitical-adaptive shipping contract (GASC) on a 1,200-node scale-free network: a dynamic geopolitical risk feed that fuses and propagates three public signals (Eqs. 1–2), a contract engine for surcharge and liability adjustment (Eqs. 3–4), risk-penalised routing (Eq. 5), a four-mode shape-following protocol (Eq. 6) and a weak budget-balance cap on cumulative surcharge. Five contracting regimes (fixed, risk-adjusted, reactive, GASC, GASC-cap) are compared under four stochastic disruption scenarios (chokepoint blockade, sanctions cascade, regional conflict, compound crisis) with 200 paired Monte Carlo runs per cell and common random numbers across regimes. Contents: the model (gasc_model.py), the full specification including the amendments adopted after the first batch (ART9_GASC_Model_Spec_v2.md), the main experiment (results/main.csv, 4,000 run–regime records), the component ablation (results/ablation.csv, 1,000 records), five sensitivity panels (results/sens_*.csv), an archived earlier batch with deterministic shocks (results/archive_v2a/), the scripts that generate every table and figure (analyze_main.py, make_tables_figs.py, make_fig1_ga.py), the tables as printed, and Figures 1–5 with the graphical abstract. No number in the paper is hand-entered; all are produced by these scripts from these files. Reproduction: python gasc_model.py pilot; python gasc_model.py main <scenario> 0 200; python gasc_model.py ablate 0 200; python gasc_model.py sens <panel> 0 100; then analyze_main.py, make_tables_figs.py and make_fig1_ga.py. Network seed 20260916; run seed = 1000 × scenario index + run. Python 3.11 with numpy, scipy, networkx, pandas, matplotlib. The full batch takes about six CPU-hours on one core; the script appends per run and resumes. This version supersedes the May 2026 draft of this record, which held a framework-only calculation package for an earlier version of the manuscript.
Author: Battal, Turgay
Institution: İskenderun Technical University
Language: English
Subfield: Strategy and Management
Domain: Social Sciences
Open Access status: green
OpenAlex ID: W7220360951