For six decades, climate scientists have meticulously built increasingly sophisticated computer models of Earth, successfully predicting the impacts of global warming. These models, pioneered by Syukuro Manabe, revolutionized our understanding of climate change, accurately forecasting rising global temperatures and associated extreme weather events. Manabe's foundational work, which earned him a Nobel Prize, relied on the assumption of 'large-scale determinism' – that large-scale climate patterns could be predicted without needing precise detail on smaller scales. This approach yielded remarkably accurate long-term predictions.
However, recent years have revealed a critical juncture. While broad predictions, like Arctic warming rates and global temperature increases, remain largely accurate, the models are faltering in their finer details. Regional predictions are proving inaccurate, with discrepancies emerging in tropical Pacific Ocean temperatures, humidity levels in southern Africa and the southwestern US, and jet stream strength. The unexpected surge in global temperatures in 2023 also highlighted a significant model shortfall.
This failure to accurately predict regional and highly detailed climate changes has prompted a 'climate crisis' of its own – a fundamental reassessment of climate modeling methodologies. Scientists like Tiffany Shaw and Bjorn Stevens argue that the long-held assumption of large-scale determinism may no longer suffice. The need for more granular predictions demands a move beyond estimations towards resolving finer-scale processes directly, demanding significantly increased computing power.
Efforts are underway to refine these models. Bjorn Stevens and his team at the Max Planck Institute are developing one-kilometer-resolution models, resolving previously uncharted mesoscale processes. Artificial intelligence is showing promise in enhancing model efficiency and automation, but a complete AI-based climate model remains elusive. However, even this technological leap may not entirely resolve the challenges, and alternative paths are being investigated.
Simultaneously, the Trump administration's assault on climate science presents a severe threat to the advancement of climate modeling. Funding cuts, staffing reductions, and the potential closure of key institutions like the Geophysical Fluid Dynamics Laboratory (GFDL) – the birthplace of many climate modeling advancements – threaten decades of progress and the careers of countless scientists. The consequences of these actions are far-reaching, potentially crippling the global effort to understand and respond to climate change. This includes the termination of crucial satellite observations and a resulting loss of valuable data used for model calibration and refinement.
The current crisis underscores a crucial lesson: The accuracy of climate models is directly tied to the societal value placed upon climate science and the continued investment in its advancement. The future success in perfecting and enhancing these models hinges on both scientific innovation and the urgent political will to address the looming climate crisis before the damage becomes irreversible.
---
Originally published at: https://www.quantamagazine.org/how-climate-scientists-saw-the-future-before-it-arrived-20250915/
