A Momentous Moment in Time

As I’m sure you’re aware, in a few days time the United Nations Framework Convention on Climate Change (UNFCCC) will meet for the twenty-first session of the Conference of Parties (COP21). Over 190 nations will gather in Paris to discuss a collective agreement on climate change, with the aim of staying within the 2^oC global warming target.  

                                            

Despite a GlobeScan poll on behalf of the BBC suggesting that public support for a global deal on climate change has declined, I hope that the outcome will reignite general public conversation and support. It’s important that everyone gets on board with this, as the conference will not only address CO₂ emissions, but also how business is conducted in terms of production and farming. Collectively, consumers can have a lot of power which needs to be utilised.

It’s likely there’s also going to be something else quite momentous going on at the same time as the COP21 summit... the peak of the 2015 El Niño event.

The scatter plot (Figure 1) and Table 1 below were produced by the International Research Institute for Climate and Society (IRI), showing forecasts made by a variety of dynamical and statistical models for Sea Surface Temperature (SST) in the Nino 3.4 region (120-170W, 5N-5S) for nine overlapping 3-month periods using observations taken in October 2015.

Figure 1: Niño 3.4 SST Anomaly (^oC) predictions from dynamical and statistical models for Nov 2015.

Source: InternationalResearch Institute

As can be seen, for all models, the highest point lies in the Nov(15) – Dec(15) – Jan(16) period. The differences between the model predictions are the result of two main factors. The first, and most obviously, being the construction of the model. Models are not able to fully capture an entire process, they are tools enabling us to analyse, predict, and learn about a process, and hence there will always be a degree of uncertainty within them. The second reason for differences is the actual uncertainty in the forecast. Forecasts made between June and December generally have a better skill* due to the reasoning that this is when El Niño events are active. As might also be expected, skills usually decrease as lead time increases.  

*A forecast “skill” is the level of accuracy a model forecasts in comparison to a reference model.

The figures in Table 1 below give average SST values for the different model types, supporting the trends seen in the scatter plot. 

Table 1: Average Niño 3.4 SST Anomaly (^oC) for given models over seasonal 3 month periods in 2015 – 2016.

Source: InternationalResearch Institute

Most of the models predict El Niño SST to slightly increase in strength until about the end of the year, and then gradually weaken through 2016. Although the El Niño will be steadily dying away, it is expected to still be a strong force through the first 4 or so months of 2016. The associated probabilities with these predictions is at least 99% up to around Feb 16, then exponentially falls away to about 50% by May – July 16.

Figure 2 below shows a longer term trend of model predictions, emphasising the strength of the El Niño which we are currently experiencing. It also shows the model forecasts for the past 21 months in addition to the Nov 2015 predictions (as seen in Figure 1). It allows us to see the evolution of model predictions as more data is gathered from the progressing conditions. Models are continually updated and re-run when new data is obtained to produce up-to-date forecasts and analysis of the processes at play.

Figure 2: Niño 3.4 SST Anomaly (^oC) predictions from dynamical and statistical models for 22 months.

Source: InternationalResearch Institute

The high SST anomalies that are being seen mean that the ocean is unable to absorb as much heat as it otherwise would during these months. This factor along with CO₂ levels breaking records and reaching a massive 400.35ppm, is what is causing the expectation that 2015 will be the hottest year on record.

Join me next time when we’ll look at what dynamical models and statistical models are, the differences between them, and why these different types of models exist for the modelling of ENSO.

El Niño in the Anthropocene

We’ve looked at the argument that El Niño is causing global warming, (and we’re hopefully all in agreement that it is obviously not) now let’s take a look to see whether global warming is affecting El Niño. 

Source: Flaming Planet by ImFayth

First, let’s just remind ourselves of a few things. Global warming is the increase in the temperature of the Earth’s surface, oceans and atmosphere. The fourth report from the Intergovernmental Panel on Climate Change (IPCC) stated, with over 90% confidence, that the majority of the observed increase in temperatures since the mid 20^th century is as result of the increased anthropogenic greenhouse gas concentrations. In a study of 11,944 peer-reviewed scientific literatures from 1991 – 2011, it was found that over 97% of climate scientists agree that anthropogenic global warming is as a cause of human activity, and not a natural occurrence (Cook et al., 2013). What the other 3% are thinking..? Well, let’s not get into that again!

This blog started with considering how much the oceans have warmed during the anthropocene, and we now know a little about the interaction between the ocean temperature and the atmosphere, (in particular El Niño episodes), so what, if any, effect is global warming having on El Niño?

One report proposed that a doubling in the occurrences of extreme El Niño events will be seen due to greenhouse warming (Cai et al., 2014). In the report, an extreme El Niño event is characterised by a massive reorganisation of atmospheric convection which leads to austral summer rainfall greater than 5mm per day over a defined area. The result partly comes from the non-linear relationship found between sea surface temperature and rainfall, and meridonial (equator to mid-latitudes) sea surface temperature gradients and rainfall. As can be seen in the plotted points in each of Figure 1 a and Figure 1 b below, if a line were to be drawn joining them they would both be a curve. There is an exponential increase and an exponential decrease in rainfall associated with an increase in SST and meridonial SST gradients, respectively.

Figure 1 a,b: Evolution and nonlinear characteristics of observed extreme El Niño events. (Niño3 area: 5◦ S–5◦ N, 150◦ W–90◦ W).  

Extreme El Niño, moderate El Niño, and La Niña and neutral events, are indicated by red, green and blue dots respectively.

Source: Cai et al., 2014

A report by Hansen et al, 2006, claims there will be an increase in intensity of El Niño events, but the effect on frequency is still unclear. Whilst Kim et al, 2014, deduced that there will be a time-varying response of ENSO to global warming, with an increasing trend in amplitude before 2040, and thereafter a decreasing trend.  

However, another report concluded that it is not yet possible to say whether there would be a change in frequency or strength of events due to the complex year to year variability of ENSO (Collins et al., 2010). Many processes play a part in determining ENSO characteristics, and the effect of global warming would increase them. However, these processes involve both positive and negative feedbacks; hence the uncertainty lies within these feedbacks and, essentially the report suggests they could cancel each other out.

It seems that it we do not currently have enough information to give a unanimous conclusion on whether global warming is going to increase the frequency and intensity of El Niño events (Guilyardi et al., 2012). But what we do know is that potentially global warming will increase the events and hence have catastrophic effects on our environment, from flooding to drought. This highlights the importance of needing to model ENSO activity. Modelling allows us to learn more about the processes by discovering underlying trends, as well as being able to simulate ‘what if?’ scenarios. What if sea surface temperatures rise by X^oC? What if sea levels rise by X inches? Considering these situations is the start of either stopping, or, if unavoidable, preparing for them.  

Planet Ocean

If you watch any film this week - let it be this. 

Planet Ocean is a powerful documentary by Yann Arthus-Bertrand which highlights the importance of the ocean in the ecosystem. 

Be absorbed into the beautiful cinematography, and learn how imperative it is that humankind lives in harmony with the oceans. 

"Nature doesn't tolerate excess"

Global Warming? Blame it on El Niño

Somehow it’s already that time of year again… November. Bam, its Christmas! Oxford Street lights have been switched on, Mariah Carey will soon be wailing out of every speaker, and what everyone is desperate to see, (don’t fret, there’s not long to wait) the 2015 John Lewis Christmas advert.

So, as tradition says, you have to buy your nearest and dearest presents. But what to get that person who has everything? I have the answer. This: 

Source: Designed by Words & Unwords for Zazzle.com   

If you’re not feeling the t-shirt, not to worry, you can get the same design on a mug, tote, or for that really special person, how about a badge?

Stylish and educational, don’t you think?  

The results of a study (McLean et al., 2009) stated that at least 68% of the variance in global tropospheric temperature anomalies (GTTA) is accounted for by the Southern Oscillation Index (SOI), i.e. ENSO. This raises to 81% accountability in the tropics.

They are some large percentages. So maybe the t-shirt isn’t being sarcastic after all?

“The close relationship between ENSO and global temperature, as described in the paper, leaves little room for any warming driven by human carbon dioxide emissions.”

This is a quote taken from the press release related to the paper by one of the authors, R. M. Carter.

Hmm, now I’m really hoping the t-shirt is, as I first thought, being sarcastic.

So how did McLean et al come to such a bold conclusion? Well, essentially by removing the long term upward trend of the global tropospheric temperature anomalies. This trend can be seen by the GTTA line in Figure 1.



Figure 1: Twelve-month running means of SOI (dark line) and MSU GTTA (light line) for the period 1980 to 2006 with major periods of volcanic activity indicated

Source: McLean et al., 2009

A mere -0.223 correlation is found between the pre-transformed variables, suggesting a very weak negative correlation. However, when both the SOI and GTTA data is transformed into time derivatives, a correlation of -0.847 is found. This is now suggesting a very strong negative correlation, and thus forms the basis for the study’s conclusion. However, these time derivatives are calculated by subtracting the 12 month moving average from the same average for data 12 months later. It is said this transformation is to remove noise, but this herein lies the problem, as it removes the linear upward trend of the temperature data. Therefore a strong correlation will be found, as it has been recognised that El Niño has a strong effect on short term global temperatures but that it cannot account for the long term trend (Santer et al. 2001, Lean and Rind 2008, Foster and Rahmstorf 2011).

In fact, in direct contrast to McLean et al., 2009, it has been suggested that El Niño has had a slight net cooling effect on global temperatures (Fawcett, 2007). This has also been noticed for the period 1979 to 2005 by Lean and Rind, 2008, and for the period 1979 to 2010 by Foster and Rahmstorf, 2011.

I hope we are now all in agreement that global warming is not the fault of El Niño? Please drop me a comment if you have any questions.

Finally, if you do decide to buy the t-shirt for someone this Christmas, hopefully you’ll now have enough knowledge to share over your turkey (or vegetarian equivalent) dinner.