Integrating swarm intelligence with CMIP climate models for ecocritical environmental analysis

Abstract

This research establishes a cohesive swarm intelligence framework used for climate simulations derived from the coupled model intercomparison project phase 6 (CMIP6), obtained from the earth system grid federation (ESGF). The study examines essential environmental variables such as near-surface air temperature (tas), sea-level pressure (psl), precipitation (pr), surface shortwave radiation (rsds), and longwave radiation (rlds). The system specifically evaluates a global mean surface temperature rise of 1.72 °C, a psl range of 980-1,030 hPa, pr anomalies averaging ±1.3 mm/day, rsds values fluctuating between 140-280 W/m², and rlds values reaching a maximum of 350 W/m² for high-emission shared socioeconomic pathways (SSP)5-8.5 scenarios. The characteristics served as inputs for decentralized particle swarm architecture aimed at identifying ecological stress signs via geographic anomaly divergence, entropy deviation, and signal intensity thresholds. The model simulated swarm behavior across temporal CMIP grids, effectively capturing changes in climatic feedback and highlighting areas of ecological instability. The swarm framework dynamically analyzes pattern-based fluctuations in model output, facilitating ecocritical evaluation of environmental risk. This hybrid method integrates physically based climate data with adaptive artificial intelligence (AI) modeling, providing an ecologically contextual understanding of earth system changes and improving predictive insights for sustainability and policy formulation.