Thread by @PatrickTBrown31

@KenCaldeira and I have a new study out, “Empirical prediction of short‐term annual global temperature variability” which presents a methodology for forecasting global average temperature variability on interannual timescales. Our live forecast is here: weatherclimatehumansystems.org/global-tempera…

The full paper is: Brown, P. T., & Caldeira, K. (2020). Empirical prediction of short‐term annual global temperature variability. Earth and Space Science, 7, e2020EA001116. doi.org/10.1029/2020EA…

Some background: Global temperature variability on sub-decadal timescales can be of substantial magnitude relative to the long-term rate of global warming.

For example, global temperature increased by ~0.42°C over the 5 years between 2011 and 2016. This is equivalent to approximately three decades worth of long‐term warming (~0.14°C per decade since the middle of the twentieth century) over a short period.

Also, subdecadal global temperature variability has an extensive spatial footprint in the sense that approximately 87% of the global surface and 99% of the global land surface exemplifies a positive linear relationship between local and global temperature deviations.

Therefore, it is not surprising that this type of short-term global average temperature variability has been shown to have considerable environmental and societal effects.

This is particularly the case because extremes like the global frequency and extent of heatwaves scale with the global average temperature:

See e.g., Fischer, E. M., & @Knutti_ETH (2015). Anthropogenic contribution to global occurrence of heavy‐precipitation and high‐temperature extremes. Nature Climate Change, 5, 560. doi.org/10.1038/nclima…

Thus, downstream impacts of anomalous heat are wide-ranging; they have been associated with phenomena as diverse as primary production, the distribution of fish, coral bleaching, agricultural output, economic growth, energy demand, human mortality & migration, & civil conflicts.

This is why we hope that accurate foreknowledge of annual global temperature anomalies will be of some value for risk mitigation purposes. See an AMS talk for subsequent work with @climateai on a more direct application to relevant impacts: ams.confex.com/ams/2020Annual…

Our methodology utilizes only global spatial patterns of annual mean surface air temperature anomalies to predict subsequent annual global temperature anomalies via partial least squares regression.

The method's skill is primarily achieved via information on the state of long‐term global warming as well as the state and recent evolution of the El Niño–Southern Oscillation and the Interdecadal Pacific Oscillation.

For the task of predicting global temperature anomalies one to four years ahead of time, we find that our method is skillful relative to simple "naive" benchmarks and it typically outperforms global climate models initialized to the observed state of the climate system.

Additionally, our method is much less computationally expensive than forecasts made with global climate models.

We will update our forecast monthly and we will also keep tabs of the chances that any given year will be a new record and/or the chances that it will breach the Paris Accord-associated limit of 1.5°C. Our live forecast is here: weatherclimatehumansystems.org/global-tempera…