Last year, 2008, was about the 9th hottest year in the instrumental record (range: 7th to 10th). It certainly wasn’t the hottest year. That record goes to either 1998 or 2005, depending on which temperature record you look at. NASA’s GISTEMP gives it (slightly) to 2005, whilst the Hadley Centre and the two satellite measures (UAH and RSS) give it to 1998. Note that only GISTEMP averages over the northern polar regions, an area which has warmed more than the planet as a whole due to the retreat of snow and ice (which makes that part of the world duller, and so able to absorb more sunlight).
A question that is often asked by the naive or disingenuous is: “If the buildup of greenhouse gases is causing global warming, then why isn’t each year hotter than the previous?” This is a simplification the more common meme: “Global warming has stopped since 1998” (I note that it ceased being “Earth hasn’t warmed in the last 10 years” when 1998 passed more than a decade into history). The latter question is addressed explicitly in this earlier post — it has to do with the amount of energy transferred from the oceans to the atmosphere, given that most (>95%, on average) of the extra energy being trapped by increased greenhouse gases in going into slowly heating the oceans and melting ice across the planet.
Climate scientists obviously recognise that the Earth’s temperate is influenced by a whole range of forcings and feedbacks. Greenhouse gases, predominantly CO2, is one forcing causing gradual, inexorable warming. Transfer of heat from the oceans to the atmosphere is another and can cause temporary warming or cooling — perhaps the most important being characterised by the El Niño / La Niña oscillation. The solar cycle, lasting an average of 11 years, is yet another. A fourth powerful but occasional influence is large volcanic eruptions which cause temporary dimming (cooling). Global climate models include all of these factors when attempting to reconstruct past temperatures, and can be used to make future predictions of climate if these forcings are specified for future scenarios.
But we don’t need a global climate model to get a rough appreciation of how these forcings affect year-to-year temperature. We can approximate them with some basic correlative analysis that are uncomplicated and straightforward to understand.
Fig. 1 shows three relevant time series. In green is the seasonal average global temperature anomaly (TA), based on the composite measure provided at WoodforTrees. This is useful as it side-steps the ‘debate’ over which of the four major temperature measures is ‘best’ – it uses all of them and corrects for different baselines. Seasons are Dec-Jan-Feb, Mar-Apr-May, Jun-Jul-Aug and Sep-Oct-Nov. In blue is the southern oscillation index (SOI), in this case inverted (so that a higher SOI generally equates with a positive forcing) and normalised to scale between 0 – 1. In red is the satellite measure of total solar irradiance (TSI; similar to sunspot number), originally reported in watts per metre squared but here also normalised to a 0-1 scale so that it can be plotted alongside temperature and the SOI. The SOI and TSI have been offset by 3 months compared to Temperature, because we would expect some lag between the forcing and the response. The period 1979 to 2009 is chosen for two reasons — it is the period covered by all four temperature measures, and it also represents a span of time sufficient to represent ‘climate’ rather than weather — 30 years.
Visually, it is clear that there is quite a good relationship between the up and down fluctuations of temperature and both SOI and TSI, although the gradual upwards trend in temperature is also apparent. We can quantify this relationship statistically. I fitted a simple linear ANCOVA model with TA as the dependent variable and Time (each season since DJF of 1978-1979), SOI and TSI as continuous predictors and volcanic forcing as a categorical predictor (VOL; a value of 1 is assigned for the few seasons following the 1982 eruption of El Chichon and the 1991 eruption of Pinatubo). I’ve ignored other complicating matters such as tropospheric aerosols. Still, the model is structurally not too bad a fit (for those familiar with the method, the residual model diagnostics confirm a reasonable conformation with assumptions, and 67% of the deviance is explained by the four predictors).
These results are useful. They show a strong positive trend in temperature (+0.0054C per season [or about 0.216C per decade]), a forcing of about +0.4C for a strong El Niño, +0.24C for the peak of solar forcing, and -0.27C for a volcanic event. With this model-based information, we can make two visual additions to the temperature plot. They are show in Fig. 2. First, the original temperature data is plotted in green. In blue is the season temperature as predicted by the ANCOVA model, based on historical forcings. The dashed trendline is the warming trend expected if all other forcings (SOI, TSI and VOL) are held at their 30-year average values.
Of relevance to the title of this post, these correlations can be used to work out (crudely!) what temperatures we might have expected over the last five years based on three different scenarios. First, if SOI, TSI and VOL were all held at their 30-year average, what would the global temperatures have been in 2004, 2005, 2006, 2007 and 2007? (AvF) Also, what if they all been at their minimum (MinF) or maximum (MaxF) 30-year values? These scenario results, along with the composite observed temperature anomalies, is shown in the table below. The last three columns represent deviations from the observed values (so, for example, in 2006 the observed anomaly was +0.69C but if all forcings had been at their 30-year average [AvF] it would have been +0.03C hotter than this, or +0.72C).
These numbers indicate that 2004 and 2007 were right on the average forcing expectation . 2005 was slightly hotter than we’d have expected (it was an moderate El Niño year and near the top of the solar cycle) and almost spot on the MaxF prediction. Conversely 2008 was quite a bit cooler and close to the MinF expectation. Why? It was a La Nina year and the sun had bottomed out in its irradiance/sunspot cycle. If 2008 had been an average year, we should have expected it to be 0.25C hotter, at around +0.75C anomaly.
To cap of this little venture into what-if land, I’ll have a bit of fun and predict what we might expect for 2009. My guess is that the SOI will be neutral (neither El Niño or La Niña), the solar cycle 24 will be at about 20% of its expected 2013 peak), and there will be no large volcanic eruptions. On this basis, 2009 should be about +0.75, or between the 3rd and 5th hottest on record. Should we get a moderate El Niño (not probable, based on current SOI) it could be as high as +0.85C and could then become the hottest on record. I think that’s less likely.
By 2013, however, we’ll be at the top of the solar cycle again, and have added about another +0.1C worth of greenhouse gas temperature forcing and +0.24 of solar forcing compared to 2008. So even if 2013 is a La Niña year, it might still be +0.85C, making it hotter than any year we’ve yet experienced. If it’s a strong El Niño in 2013, it could be +1.2C, putting it way out ahead of 1998 on any metric. Such is the difference between the short-term effect of non-trending forcings (SOI and TSI) and that inexorable warming push the climate system is getting from ongoing accumulation of heat-trapping greenhouse gases.