Predicting Climate Change: The Limits of Current Models
For over fifty years, scientists have been striving to forecast Earth’s climate, yet uncertainties remain as some regions warm more intensely than anticipated. Unexpected extreme weather events are challenging our predictive capabilities, suggesting that the future might be far more unpredictable than existing models suggest.
The urgency for precise climate predictions is growing as the planet heats up. City planners, public health officials, and the global economy’s stakeholders are eager to understand regional climate changes. However, the geographic detail required surpasses what current climate models can deliver, and the computational power needed for more precise forecasts is yet unavailable.
Current climate models are adept at painting a broad picture of Earth’s climate future. Yet, scientists are observing discrepancies between predictions and actual outcomes as warming advances. According to research by Columbia University’s Kai Kornhuber, unexpected heat anomalies have emerged across various continents. Upcoming studies by Dartmouth’s Alexander Gottlieb and Justin Mankin reveal that temperature records in regions housing a third of humanity exceed model forecasts. A mid-2023 global temperature surge remains largely unexplained, concerning NASA’s Gavin Schmidt, although he acknowledges it is not entirely surprising.
“From the 1970s on, people have understood that all models are wrong,” Schmidt notes. “But we’ve been working to make them more useful.” This reflects an ongoing scientific process, though the immediate need for precise climate data for decision-making highlights the urgency of the situation.
The complexity of Earth’s systems, with intricate feedback loops between temperature, land, air, and water, complicates climate modeling. Each region’s unique characteristics, coupled with natural variability and human-induced warming, further challenge these models. Some systems, like cloud formation, remain poorly understood, impacting climate predictions. As Robert Rohde of Berkeley Earth explains, models must approximate these systems due to resolution limitations, which is why phenomena like tornadoes remain elusive in global models.
Scaling models to match human experiences is currently unfeasible, as it demands immense computational resources. Modeling Earth at a finer resolution would require dramatically more computing power, according to Schmidt. Despite these challenges, combining global models with regional data can offer valuable insights for local decision-making, although this requires expert guidance. Rohde points out, “We are asking a lot of the models. More than we have in the past.”
As models confront the realities of rapid climate change, their limitations become apparent. Initially, models successfully projected global temperature changes due to rising greenhouse gas emissions, but they struggle with unforeseen secondary effects. Certain variables, like carbon absorption by trees and land, are not fully integrated into models. Recent studies indicate that these natural carbon sinks are diminishing, altering emissions dynamics.
Biases in models present another challenge, with some overestimating risks and others underestimating them. While some models project excessive warming, recent findings on temperature extremes suggest underestimations in several regions. Rohde emphasizes that underestimating risk poses significant dangers.
For Kornhuber, the apparent underestimation of climate risks in various areas signals a concerning limitation in our understanding of Earth’s systems. “It should be worrying that we are now moving into a world where we’ve kind of reached the limit of our physical understanding of the Earth system,” he says.
As models grapple with capturing the current state of the planet, the world continues to evolve beyond familiar conditions. Given ample time, science might develop models that better reflect ongoing changes, but by then, it might be too late for effective action. Despite decades of modeling efforts, our best tools offer only estimates of the future. Rohde concludes, “At the end of the day, we are all making estimates of what’s coming. And there is no magic crystal ball to tell us the absolute truth.” What remains is a partial picture that hints at an unprecedented future.
Original Story at www.theatlantic.com