El Nino/Southern Oscillation Dynamics
I have been studying the dynamics of the ENSO system using coupled climate models ever since I began graduate school. My work in this area focuses in large part on the mechanisms which set the overall strength of ENSO variability, how well we can detectchanges in ENSO strength given the large internal variability in the system, and what role external influences on the climate (both natural and otherwise) may play in altering the characteristics of El Nino and La Nina events.
Figure 7 from Stevenson et al. (2017), Climate Dynamics
Composite evolution of EP El Niño events in the LME ensembles, shown using a Hovmoeller diagram of SSTA over 2∘S–2∘N. a Shows EP El Niños in the 850 control simulation; subsequent left-hand panels (b, d, f, h, j, l, n) show differences between the pre-industrial portions of the forced LME ensembles relative to the control, and right-hand panels (c, e, g, i, m, o) show differences between the twentieth century and pre-industrial portions of individual LME ensembles. Stippling indicates that a Wilcoxon rank-sum test at that grid point resulted in SST anomalies indistinguishable from one another at 90% significance. f Is blank since the 850 control is used as the pre-industrial portion of the ozone/aerosol-only (O3AER) ensemble.
ENSO 'Diversity'
In the past few years, the research community has begun to focus on "ENSO diversity", or understanding the differences between the properties of given El Nino/La Nina events. Even over the past 30 years, we have seen a variety of different 'flavors' of El Nino events: for instance, the very weak, debatably "failed" El Nino of 2014-15, then the major "Godzilla" event of 2015-16.
The development of the Community Earth System Model (CESM) Last Millennium Ensemble has allowed us to look at how natural and anthropogenic influences have affected ENSO diversity in more detail than ever before. The LME is a suite of global simulations with the CESM, covering the 850-2005 period, and includes subsets of simulations which are forced with only a single external influence (e.g. solar irradiance changes, greenhouse gas emissions, volcanic eruptions, etc.)
Our recent Climate Dynamics ENSO diversity special collection manuscript (Stevenson et al. 2017) describes how ENSO diversity responds to all of these forcing factors individually: we find that over the 20th century, greenhouse gas effects on El Nino development are largely offset by aerosol (air pollution) emissions. However, changes to land use and land cover also play a large role in setting El Nino strength: this is critically important for understanding climate model projections, since all 21st century scenarios include changing land use as a component of forcing.
Large Ensemble Analyses
Changes to ENSO behavior are one aspect of future climate that models still cannot tell us about with much confidence: for the past 10 years, models have been roughly evenly split on the question, with half projecting an increase in ENSO variability and half a decrease. New, large ensembles with state-of-the-art climate models still do not provide any more agreement, but their large size allows us to diagnose the physical mechanisms for inter-model differences without worrying about complications from internal variability.
I am currently working on a diagnosis of two 30-member ensembles: the CESM Large Ensemble and an ensemble run with the NOAA Geophysical Fluid Dynamics Laboratory's ESM2M.
Past Projects: ENSO and Climate Change
Much of my PhD thesis focused on understanding the response of ENSO variability to 21st century climate change. Topics covered included:
- Construction of a new statistical forecasting model for ENSO using a generalized linear modeling framework; this model performed comparably to state-of-the-art operational forecasting systems
(Stevenson et al. 2013)
- Development of a new, wavelet-based toolkit for assessing the statistical significance of differences in ENSO behavior across two time series (Stevenson et al. 2010)
- Analysis of projected 21st century changes to ENSO and its atmospheric teleconnections in simulations with models from NCAR and other modeling centers conducted for the fifth Coupled Model Intercomparison Project (CMIP5)
(Stevenson et al. 2012a, Stevenson et al. 2012b, Stevenson 2012)