State of the climate: Warm start to 2018 despite La Niña conditions

The first three months of 2018 have been between the fourth and sixth warmest first quarter on record since 1880. This is despite a modest La Niña event leading to a relatively cooler start to the year, compared to recent record warmth.

Overall, 2018 is on track to be the fourth warmest year on record after 2016, 2017, and 2015, but depending on how the rest of the year shapes up it could be as high as the second warmest, or as low as the 12th warmest.

In the latest quarterly “state of the climate” report, Carbon Brief reviews the various metrics for 2018 to date and looks ahead at what the rest of the year might bring.

A cooler start to 2018

Global surface temperatures have warmed about 1.1C since pre-industrial times and 0.8C since the 1970s. Despite this long-term warming trend, each year does not necessarily set a new record for warmth. Short-term variations in the Earth’s climate associated with El Niño and La Niña events help make some years warmer and some cooler.

For example, 2016 was exceptionally warm with help from a strong El Niño event. In contrast, 2017 was largely “neutral”, having no strong El Niño or La Niña conditions. The start of 2018, on the other hand, has been cooled by a modest La Niña event.

While scientists predict that the La Niña will fade and switch back to El Niño conditions in the coming months, it will only have a modest effect on 2018 as surface temperatures tend to lag El Niño and La Niña conditions by a few months.

Temperatures in the first quarter of 2018 have been between the fourth and sixth warmest on record since 1880. The figure below shows global temperatures from six different scientific groups between 1970 and 2017, along with the Q1 2018 temperatures as dots.

Annual global mean surface temperatures from NASA GISTemp, NOAA GlobalTemp, Hadley/UEA HadCRUT4, Berkeley Earth, Cowtan and Way, and Copernicus/ECMWF, along with 2018 temperatures to-date (January-March). Anomalies plotted with respect to a 1979-2016 baseline. Temperatures only available to the end of February in the Hadley/UEA, Cowtan and Way, and Berkeley Earth datasets. Chart by Carbon Brief using Highcharts.

Looking at monthly rather than annual temperatures, the El Niño event in 2015/2016 can clearly be seen, through the warming bump in early 2017 does not appear to be associated with an El Niño.

Temperatures in the first three months of 2018 have been largely similar to the second half of 2017, though there has been a particularly wide variation in temperatures reported by different groups resulting from differences in how much of the Arctic and Antarctic that they cover.

Monthly global mean surface temperatures from NASA GISTemp, NOAA GlobalTemp, Hadley/UEA HadCRUT4, Berkeley Earth, Cowtan and Way, and Copernicus/ECMWF. Anomalies plotted with respect to a 1979-2016 baseline. Chart by Carbon Brief using Highcharts.

The first three months of 2018 can give some sense of what to expect for the entire year. By looking at the relationship between the first three months and the annual temperatures for every year since 1950, as well as current and forecast El Niño/La Niña conditions, Carbon Brief has created a projection of what the final global average temperature for 2018 may turn out to be. Our analysis also estimates the large uncertainty in 2018 outcomes, given that temperatures from only three months of the year are available so far.

The figure below shows the results of Carbon Brief’s analysis using the NASA GISTemp dataset. The yellow dot shows the average 2018 temperature over the first three months, while the red square shows the most likely 2018 annual temperature estimate. The error bars surrounding it indicate the 95% confidence interval – namely, there’s a 95% likelihood that the 2018 global average temperature will fall in this range.

Annual global average surface temperature anomalies from NASA plotted with respect to a 1979-2016 baseline. To-date 2018 values include January-March. Estimated 2018 annual value based on relationship between January-March mean, ENSO index and projections, and annual temperatures since 1950. Chart by Carbon Brief using Highcharts.

Based on the NASA dataset, 2018 is likely to be the fourth warmest year on record after 2016, 2017 and 2015. However, this early in the year, the uncertainty around this estimate is quite large; 2018 could reasonably end up anywhere between second and eighth warmest on record. NASA estimates temperatures over polar regions where observations are a bit more sparse, and generally reports temperatures close to other groups like Cowtan and Way and Berkeley Earth that employ similar interpolation approaches. The Copernicus/ECMWF record also shows more warmth in recent months.

Looking at the temperature record from the US National Oceanic and Atmospheric Administration (NOAA), which includes less of the Arctic and Antarctic regions in their global temperature estimate, Carbon Brief’s estimate of likely 2018 temperatures is a bit lower. While the NOAA dataset similarly suggests that 2018 is most likely to be the fourth warmest year on record, it also has a wide range from as high as third to as low as 12th. NOAA’s record has generally been quite similar to that of the UK’s Met Office Hadley Centre in recent years.

Carbon Brief’s projection is consistent with the Met Office’s 2018 forecast, published in December 2017. They estimated that 2018 is most likely to be the fourth warmest year on record, but could be anywhere between the second and fifth warmest.

Annual global average surface temperature anomalies from NOAA plotted with respect to a 1979-2016 baseline. To-date 2018 values include January-March. Estimated 2018 annual value based on relationship between January-March mean, ENSO index and projections, and annual temperatures since 1950. Chart by Carbon Brief using Highcharts.

Recent temperatures can also be compared to projections from the climate models included in the latest Intergovernmental Panel on Climate Change (IPCC) report. While climate models show their own El Niño- and La Niña-like behaviour, it does not necessarily occur at the same time in models as it does in the real world. The figure below shows a running average of the last 12-monthly temperatures from 1979 through present from both observations and climate models, with the average of all models shown in black and the 95% range of the models shown in grey.

Running 12-month average global average surface temperatures from CMIP5 models and observations between 1970 and 2020. Models use RCP4.5 forcings after 2005. They include sea surface temperatures over oceans and surface air temperatures over land to match what is measured by observations. Anomalies plotted with respect to a 1979-2000 baseline. Chart by Carbon Brief using Highcharts.

During the 2015/2016 El Niño event, temperatures were a bit above the average climate model projection; now that it has ended they are slightly below. But all temperatures observed in recent years have been well within the range of model projections.

Low sea ice at both polls

Despite the somewhat cooler start to 2018, sea ice has been at record lows in the Arctic and quite low in the Antarctic. The figure below shows both Arctic and Antarctic sea ice in 2018 (solid red and blue lines), the historical range in the record between 1979 and 2010 (shaded areas), and the record lows (dotted black line). Unlike global temperature records, sea ice data is collected and updated on a daily basis, allowing sea ice extent to be viewed through to the end of April.

Arctic and Antarctic daily sea ice extent from the US National Snow and Ice Data Center. The bold lines show daily 2018 values, the shaded area indicates the two standard deviation range in historical values between 1979 and 2010, and the dotted black lines show the record minima for each pole. Chart by Carbon Brief using Highcharts.

For the first four months of 2018, Arctic sea ice extent set a new record low nearly every day. However, it is too early to tell how low the summer minimum in the Arctic will be, as the size of the summer low will also depend on Arctic weather during the coming months.

Antarctic sea ice extent set new record lows for a brief period in February and has generally been on the low-end of normal for the first four months of 2018.

To look at the long-term changes in both Arctic and Antarctic sea ice, Carbon Brief converted the data to “anomalies”, or a change relative to the 1979 and 2010 average. The figure below shows Arctic sea ice extent anomalies for each day since 1978 as blue dots, along with the trend shown in yellow.

Arctic daily sea ice extent from the US National Snow and Ice Data Center from 1978 to present. Yellow line shows a smoothed (LOWESS) fit to the data to show the long-term trend. Chart by Carbon Brief using Highcharts.

While there is some year-to-year variability in Arctic sea ice extent, the long-term downward trend is clear and is directly related to rapid warming in the region. Most years over the past decade have at least one million square kilometres less sea ice than back in the 1980s.

Summer sea ice has declined faster than winter, leading to an apparent increase in variability in the year-to-year anomalies. While summer lows have yet to fall below their 2012 record, there has still been a clear reduction in sea ice during most months in recent years, as can be seen from the smoothed fit (yellow line). Arctic sea ice volume has also declined dramatically between 1979 and present.

The changes in sea ice in the Antarctic are much less straightforward. While the last two years have seen record lows, there is little indication of a long-term trend related to warming – and Antarctic sea ice actually saw record highs as recently as 2014 and 2015. Scientists hoping to solve this puzzle say both natural and human-caused influences could be involved.

Antarctic daily sea ice extent from the US National Snow and Ice Data Center from 1978 to present. Yellow line shows a smoothed (LOWESS) fit to the data. Chart by Carbon Brief using Highcharts.

Global sea ice shows a combination of Arctic and Antarctic influences, with a modest decrease over the past 40 years coupled with a dramatic drop after 2015. However, despite a tendency to focus on recent record low global sea ice extent, it is often more useful to look at the Arctic and Antarctic separately as markedly different factors influence each region.

Global (Arctic + Antarctic) daily sea ice extent from the US National Snow and Ice Data Center from 1978 to present. Yellow line shows a smoothed (LOWESS) fit to the data. Chart by Carbon Brief using Highcharts.