Should We Expect Each Year in the Next Decade (2019-2028) to be Ranked Among the Top 10 Warmest Years Globally?

A new operational algorithm is introduced for projecting the annual rankings of NOAA’s global surface temperature over the next decade, including the “Monitoring Year” ranking (i.e., the end-of-year ranking). Two methods considered yielded comparable performance statistics over a 1999-2018 verification period: (1) a simple autoregression of de-trended monthly anomalies superimposed on an extrapolated trend and (2) a variant of (1) which includes a simple regression-based El Niño – Southern Oscillation adjustment. For simplicity and operational expediency, we propose use of method (1) as part of NOAA’s climate monitoring suite of monthly products. Extrapolating the linear trend reduced the mean absolute rank error by ~19% (~65%) at Year 5 (10). Overall, the algorithm yielded mean absolute rank errors that ranged from 0.4 (at 1-3 month leads) to 2.2 (at a 10-year lead), suggesting a high level of predictability and likelihood that most, if not all, of the next ten years (2019-2028) will rank in the top ten over the applicable period of record, at least initially. Given this absolute warmth, we also introduce a simple new annual global surface temperature score to distinguish between warm (e.g., 1998 and 2016) and cold years (e.g., 2008 and 2011) relative to long-term trends.