Donald Trump’s surprise November election as the U.S.’s forty-fifth President isn’t the only development that made 2016 a landmark year for Earth’s climate.

Even more importantly, over the long run in any event, may be the news the climate itself made over that 12-month period.

For one thing, it was the warmest year globally since records began in 1850, and probably the warmest year for at least a few thousand years. It was also the first full year the planet had more than 400 parts per million (ppm) CO2 in the atmosphere, up from 280 parts per million in pre-industrial times. Current atmospheric CO2 concentrations are probably the highest that Earth has experienced over at least the past three-million years.

”State Click To Tweet

Twenty-sixteen also brought forth record-low average Arctic sea ice for much of the year, and the second lowest minimum sea ice cover reported. Warming continued apace, consistent with projections of climate models. And it’s the year in which any credible sign of a pause or “hiatus” in the warming of Earth’s surface through present day was thoroughly refuted.

Carbon Dioxide

For the past million years, atmospheric CO2 has ranged from around 280 ppm during warm interglacial periods to as low as 180 ppm during cold ice ages. It had likely remained below 400 ppm for at least the past three million years.

Since pre-industrial times, atmospheric CO2 has increased steadily from 280 ppm to around 404 ppm. Concentrations passed 400 ppm in 2014, but 2016 is the first year it remained above 400 ppm for the entire year. Barring some breakthrough by which CO2 can be effectively sucked-out of the atmosphere, it will remain above 400 ppm at least through the lifetimes of all humans now alive.

Figure 1
Figure 1: Atmospheric CO2 concentrations (ppm) in the top panel, year-over-year increase in atmospheric CO2 (ppm) in the middle panel, and man-made emissions of CO2 from burning fossil fuels and cement production (in gigatons carbon) in the bottom panel. Data from the Global Carbon Project.

Final numbers for 2016 have not yet been released, but preliminary estimates of global CO2 emissions in 2016 suggest that emissions have remained flat, even as atmospheric concentrations increased at a record rate between 2015 and 2016. This seeming discrepancy is mostly the result of the large El Niño event, which tends to lead to a larger portion of emissions remaining in the atmosphere (a similar phenomenon can be seen back in the large El Niño of 1998).


Six different groups around the world provide estimates of global surface temperatures: NASA, NOAA, the UK’s Hadley Centre/CRU, the Japanese Meteorological Agency (JMA), Berkeley Earth, and the researchers Cowtan and Way. All agree that 2016 was the warmest year on record, though they differ a bit on the margin. All also agree that 15 of the past 16 years were the warmest ever recorded. Global temperatures from these groups (and also of an “uncorrected” record using only raw land and ocean data) are shown in the figure below:

Figure 2
Figure 2: Global annual surface temperatures from 1880 through 2016 for the major groups, as well as an uncorrected series based on raw GHCNv4 and ICOADS data. HadCRUT refers to the Hadley Centre/CRU dataset.

Zooming in on the period after 1970, one sees a record of largely unabated warming, with temperatures increasing steadily accompanied by some short-term variability driven by El Niño and La Niña events, and also by major volcanic eruptions like Pinatubo in 1992.

Figure 3
Figure 3: Global annual surface temperatures from 1970 through 2016.

There is no longer any evidence of a pause in warming through present in any of the land surface temperature records. Indeed, the rate of warming after 1998 (the so-called “hiatus” period) is now actually faster than in the period before 1998, as shown in the figure below using NASA’s GISTemp data.

Figure 4
Figure 4: Global annual surface temperatures from 1970 through 2016 for NASA GISTemp, with trends from 1970-2016, 1970-1997, and 1998-2016 shown.

There of course are still some interesting scientific questions regarding short-term variability in the period between 1998 and 2012, but it now is clear that the rate of warming has in no way slowed down. And while the recent temperatures are high in part because of a large El Niño event, the inclusion of an equally large El Niño event at the start of the “hiatus” period (1998) means that trends between the two should be relatively unbiased by either.

Figure 5
Figure 5: Record margin for 2016 temperatures relative to 2015 for surface observations and 1998 for satellite observations.

Finally, different series have shown differing degrees of record temperatures in 2016. Surface temperature data from NOAA and Hadley showed a relatively modest increase compared with 2015, while data from Cowtan and Way, NASA, and Berkeley Earth showed a much larger and more robust record.

These differences in surface records largely come down to how Arctic coverage is handled. Series like NOAA and Hadley do not include much of the Arctic, which experienced unprecedented warmth in the last three months of 2016. Cowtan and Way, NASA, and Berkeley Earth, by contrast, all include the full Arctic, and thus show more warming in 2016.

Satellite-based measurements of lower tropospheric temperatures (about 2 to 5 miles up in the atmosphere) also vary in how much of a record 2016 was. In RSS version 4 and UAH version 5.6, 2016 was much warmer than the next warmest year in the satellite record, 1998. In the new UAH beta version 6 and the old RSS version 3, 2016 only barely edged out 1998. In general, there are much larger differences between satellite records than surface records, reflecting the bigger impact that adjustments for factors like orbital decay and satellite transitions make for the resulting temperature records (compared to the relatively small adjustments made to surface records by different groups). Tropospheric temperatures measured by satellites also show a much larger response to El Niño events, which is why 1998 was so much warmer in satellite records than surface records.

Sea Ice

2016 was a year of unusually low sea ice extent both in the Arctic and Antarctic regions. Record low monthly extents were set in the Arctic in January, February, April, May, June, October, and November; and in the Antarctic in November and December. The minimum Arctic sea ice extent (in September 2016) was tied with 2007 for the second lowest on record, behind only 2012.

Figure 6
Figure 6: Daily sea ice extent for 2016 and the first month of 2017 (solid lines), compared to the 1978-2010 average (dashed lines) and two-sigma range (shaded areas).

Sea ice extent for both Arctic and Antarctic regions is shown in the figure above. For the Arctic, sea ice extent was more than two standard deviations below the 1978-2010 averages for virtually the entire year. Antarctic sea ice was relatively normal until the last three months of the year, when it fell to record lows. Arctic sea ice is continuing a long-term decline driven by warming temperatures. Antarctic sea ice, on the other hand, likely reflects short-term variability and may be unrelated to temperatures, as there has been no long-term declining trend in Antarctic sea ice. Either way, both Arctic and Antarctic sea ice will be fascinating to watch in 2017.

Zeke is an energy systems analyst and environmental economist with a strong interest in conservation and efficiency. He was previously the chief scientist at C3, an energy management and efficiency company,...