This research article from Science of Climate Change investigates whether climate models are truly useful for informing policy decisions, emphasizing the importance of out-of-sample predictive accuracy over simply fitting historical data.

Kesten C. Green and Willie Soon compare climate models used by the IPCC, which focus on human influence (Anthro), volcanic activity (Volcanic), and a specific measure of solar activity (Solar IPCC), against alternative models incorporating independent measurements of solar activity.

Their findings suggest that models including independent solar variables consistently provide more accurate forecasts than the IPCC's anthropogenic models, especially when predicting temperatures in rural areas less affected by urban heat islands.

The study concludes that the IPCC's models lack predictive validity and are unreliable for forecasting future temperatures, advocating for the use of out-of-sample forecast errors to evaluate model effectiveness.

Summary and podcast generated by NotebookLM.

Get full access to Dead Man Talking at metatron.substack.com/subscribe