Each El Niño event looks different in terms of their magnitudes, spatial structures, and temporal evolutions. One important dynamical factor that leads to this El Niño diversity is stochastic atmospheric forcing such as WWBs that episodically occur in the tropical Pacific (Fig. 1a). WWBs are often associated with twin tropical cyclones or the Madden-Julian Oscillation, and their occurrence results in the relaxation of the easterly trade winds that can last from days to a few weeks. Our recent studies suggest that WWBs have strong impact on the development, diversity, and predictability of El Niño events, and this impact is dependent on the ocean initial state when the WWBs occur (Hu et al., 2014; Fedorov et al., 2015). EWBs (Fig. 1b), on the other hand, can effectively stall El Niño development, exemplified by the recent failed El Niño of 2014-2015 (Hu and Fedorov, 2016).
Fig. 1: Examples of (a) WWB and (b) EWB. Anomalies in sea surface temperature (colors; °C) and surface wind (vectors; m/s)
The recent El Niño development in 2014-2016 nicely illustrates how an interplay of WWBs and EWBs affects El Niño development and predictability. In the early months of both 2014 and 2015, there were clear indications pointing to the possibility of a strong El Niño, including a heat-recharged equatorial ocean and a series of strong WWBs. However, in 2014 the warm event was stalled by an exceptionally strong easterly wind burst and did not exceed the formal threshold for El Niño (Hu and Fedorov, 2016; Fig. 2a). Nevertheless, the failed 2014 event preconditioned for El Niño development in the following year, and the equatorial warming in 2015 developed into an extreme event (Hu and Fedorov, 2017a; Fig. 2b). These conclusions are supported by satellite-based and other observations, and by large-ensemble coupled simulations with superimposed wind bursts.
Fig. 2: (a) The failed El Niño of 2014 and (b) the extreme El Niño of 2015. Anomalies in sea surface temperature (colors; °C) and surface wind (vectors; m/s)
Slower rates of increase in global mean surface temperature (GMST) after 1998, dubbed “global warming hiatus”, recently gave way to a rapid GMST increase. This temperature rise coincided with the persistent warm conditions in the equatorial Pacific between 2014-2016, which has led to a large amount of heat release into the tropical atmosphere. Recently we construct a simple, physically-based model of GMST variations that incorporates greenhouse gas emissions, ENSO, and stratospheric sulfate aerosols, and this model closely reproduces GMST changes since 1880, the global warming hiatus, and the subsequent temperature rise (Hu and Fedorov, 2017b; Fig. 3). Our results confirm that weak El Niño activity was the major cause of the hiatus, while the rapid temperature rise is due to atmospheric heat release during 2014-2016 El Niño conditions concurrent with the continuing global warming trend.
Fig. 3: GMST variations estimated from the observations (black) and computed from the simple model (red).
The “elevated heating” effect refers to the phenomenon that temperature at a given level of the atmosphere is expected to be higher over an elevated surface than over a non-elevated surface. This phenomenon has been known for long, but its physics still remains unclear. The Tibetan Plateau, about 4 km higher than the adjacent Indian subcontinent, was long thought to act as an elevated heat source that drives the South Asian summer monsoon. However, this idea has been recently challenged by the observations that atmospheric energy content peaks south of the plateau, rather than over it. So what controls the strength of elevated heating in RCE? How will it change with climates? Why is the expected elevated heating absent over Tibet? These are the questions I want to address in the project with Prof. William Boos.