A new study by Atwood et al (2021) published in the journal of Paleoceanography and Paleoclimatology found there’s “poor agreement” between precipitation reconstructions and model simulations over the past 2000 years. This means future projections made by current models are unreliable.
Die kalte Sonne here reports on a team of scientists who examined 67 tropical hydroclimate records from 55 sites around the world. The key points:
Models and reconstructions don’t agree
These comprehensive reconstructions show that from 800 to 1000 CE there was a pronounced drying event relative from the eastern Pacific and parts of Mesoamerica.
Also the period “1400–1700 CE is marked by pronounced hydroclimate changes across the tropics, including dry and/or isotopically enriched conditions in South and East Asia, wet and/or isotopically depleted conditions in the central Andes and southern Amazon in South America, and fresher and/or isotopically depleted conditions in the Maritime Continent.”
The study’s abstract also notes how there’s a glaring disagreement between the simulations done by models and what the reconstructions show: “We find notable dissimilarities between the regional hydroclimate changes and global-scale and hemispheric-scale temperature reconstructions, indicating that more work needs to be done to understand the mechanisms of the widespread tropical hydroclimate changes during the LIA.”
The problem, the authors say, is that “climate model simulations exhibit weak forced long-term tropical rainfall changes over the last millennium” and the reconstructions tell a different story.
Climate forcings not understood
The inability for the models to perform adequately are likely due to the “inaccurate” use of climate forcings and poor estimation of internal climate variability. Another remote possibility is that the proxy records are simply being misunderstood.
Models not reliable for forecasting
Die kalte Sonne comments: “Because the models do not correctly reproduce the past, their forecasting quality for the future must be taken very critically. Visions of future horror droughts or Biblical floods should be ignored until the models are able to successfully reproduce the documented global precipitation history.”