This work sits apart from neighboring studies: the closest in spirit deal with El Niño prediction, Antarctica's response to carbon dioxide, and solar influence on ice ages. Unlike them, this paper doesn't hunt for external drivers or long-period cycles. Instead, it tests the hypothesis that the twentieth-century warming trend itself wasn't smooth but stepped. This continues a line of research on regime shifts in the climate system that began back in the 1990s, but with a stricter definition: a shift counts as significant only if it can't be explained by known within-regime variability, such as ENSO or the PDO.
The authors used monthly sea surface temperature anomalies from two datasets: HadSST2 for the period up to the early 1980s and Reynolds v2 for later times, considering satellite data more accurate in spatial coverage. For each latitude band (tropics 30°S–30°N and northern mid-latitudes 30°N–60°N), they built a multiple linear regression where predictors were the El Niño index Nino34 with a one-month lag or the Pacific Decadal Oscillation index with no lag, plus a step function for climate regime, equal to -1 before 1926, 0 between 1926 and 1987, and +1 after 1987. Coefficients were fitted by standard least squares in Excel. Shift dates were determined by a grid search: the chosen time was such that the reconstructed curve showed no visible breaks against observations.
For the tropics, the two-factor model (ENSO plus climate regime) explains 86% of the variance in monthly temperature anomalies over 1900–2012. For northern mid-latitudes, a similar model with the PDO instead of ENSO gives a correlation of 0.81. The estimated size of each of the two jumps was 0.28°C for the tropics and 0.36°C for northern mid-latitudes; when the shifts are treated separately, their amplitudes come out close: 0.26 and 0.28°C for the tropics, 0.38 and 0.34°C for northern mid-latitudes. Robustness checks showed that coefficients obtained from 1910–1940 data alone reproduce the full record with correlations of 0.86 for the tropics and 0.81 for northern mid-latitudes, indicating the stability of the found relationships.
Unlike neighboring works that seek external forcings or predict individual phenomena, this paper offers a minimalist description of the warming phenomenon itself: not a continuous trend but two sharp shifts against a backdrop of known natural variability. This approach changes the view on causality: perhaps a significant part of twentieth-century warming isn't tied to the gradual buildup of greenhouse gases but to rare reorganizations of the climate system, whose mechanisms have yet to be understood. The work also shows that complex models aren't always necessary: two factors and a step function give nearly the same explanatory power as far more detailed reconstructions.
The topic could develop toward finding physical mechanisms for the detected shifts. If they can be linked to specific processes, for instance a reorganization of ocean circulation or threshold effects in the atmosphere, that would allow forecasting the next such event. Another direction is testing the hypothesis on independent data: paleoclimate reconstructions, observational records from other periods, or other latitude bands, including the Southern Hemisphere.
The results could affect the interpretation of climate models and approaches to attributing observed warming. They are also useful for ecologists and biologists studying regime shifts in ecosystems, since they provide a simple statistical tool for picking out step changes from noise.
The immediate next step is to reproduce the analysis on newer data (post-2012) and on independent temperature datasets to confirm the stability of the shift dates and amplitudes. It's also important to check whether the step structure persists when using other indices of natural variability, such as the Atlantic Multidecadal Oscillation index.
This work directly addresses an unresolved problem in climate science: the relative roles of internal climate variability and external forcings in the observed warming. Finding sharp shifts not explained by known oscillations raises the question of unknown threshold mechanisms, which is a central theme in modern climate science.
The 1976/77 shift, often considered the largest regime shift of the twentieth century, isn't recognized as such in this study: it's almost entirely explained by natural El Niño variability. The 1987/88 shift, in contrast, stands out precisely because it can't be reduced to known oscillations.