Claims that terrestrial temperature records are unreliable and that satellite records are the only basis for assessing temperature trends have been made. Also, claims that the satellite records show no global warming since 1979. This page contains our response to such claims. A paper reviewing the statistics of global mean temperature trends is described; this adds weight to our own conclusions.
2016 Temperature Records and "The Pause"
Gavin's comments, 19 January 2017
To nobody's surprise, all the surface datasets showed 2016 to be the warmest year on record.
Barely more surprising is that all of the tropospheric satellite datasets and radiosonde data also have 2016 as the warmest year.
Coming as this does after the record warm 2015, and (slightly less definitively) record warm 2014, the three records in row might get you to sit up and pay attention.
There a few more technical issues that are worth mentioning here.
Impact of ENSO
The contribution of El Niño to recent years' anomalies in the GISTEMP data set are ~0.05ºC (2015) and ~0.12ºC (2016), and that means the records would still have been set even with no ENSO variability.
I calculated these values using a regression of the interannual variability in the annual mean to the Feb-Mar MEI index. This has (just) the maximum correlation to the annual means (r=0.66). The impact of ENSO on other indices is similar, but does vary - the datasets that don't interpolate to the Arctic, in recent years at least, have a slightly stronger ENSO signal, as do the satellite tropospheric records. Doing the same procedure with the HadCRUT4 data, does change the ordering - with 2015 staying as the record year, but using the Cowtan and Way extension, the results are the same as with GISTEMP. Which brings us to another key point…
Impact of the Arctic
It's perhaps not obvious in the first figure, but the magnitude of the record in 2016 is much larger in GISTEMP and Cowtan &Way (and in the reanalyzes), than it is in HadCRUT4, NCEI and JMA. This is in large part due to the treatment of the Arctic. The latter 3 records all 'conservatively' don't include areas where there aren't direct observations in their global means. This is equivalent to assuming that the missing areas are, on average, warming at the same rate as the global mean. However, this has not been a good assumption for a couple of decades. Arctic anomalies this year were close to 4ºC above the late 19th Century, over 3 times as big an anomaly as the global mean.
This divergence between the 'global' averages wouldn't matter if all comparisons were done against masked model output, but this is often skipped over for simplicity. I personally think that both HadCRUT4 and NCEI should start producing a 'filled' dataset using the best of the techniques currently available so that we can move on from this particular issue.
Do I have to mention the 'pause'?
Apparently yes. The last three years have demonstrated abundantly clearly that there is no change in the long-term trends since 1998. A prediction from 1997 merely continuing the linear trends would significantly under-predict the last two years.
The difference isn't yet sufficient to state that the trends are accelerating, but that might not be too far off. Does this mean that people can't analyze interannual or interdecadal variations? Of course not, but it should serve as a reminder that short-term variations should not be conflated with long term trends. One is not predictive of the other.
JACK: Anywhere on the global record a sequence of 20 years may be selected which either shows a negative slope or a positive one. Cherry-pickers note. BUT overall there is a positive temperature trend beginning in 1850! Stop sniping.
Here's the HADCRUT4 record of global temperature anomalies since 1850. This precedes the discussion of the argument that no change in global temperature has occurred since 1979 and the claims by Nigel Lawson's group that there have been no changes since 1998 and that this disproves the physics of the greenhouse effect.
The above graph shows the general rise in temperature since 1850. Also reasonably obvious are the decreases from 1875 to 1910 and from 1940 to 1976. These are due to the negative phases of the Pacific Decadal Oscillation now occurring since 1998 and causing the present almost zero trend until the recent data for 2015 that show some further warming.
No warming since 1979?!
This title represents some people's opinions of the temperature records since 1979 when satellite measurements began. The data from the horse's mouth [UAH, Spencer and Christy] are shown in Figure 1.
Figure 1 Lower troposphere temperatures since 1979 [from www.ddroyspencer.com]
Roy has chosen to present the trend of the data by a sine curve or a polynomial curve with the first half of the data having a general temperature lower than the second half, divided by the extraordinary El Niño of 1998. Over a limited period of 32 years since the satellite records have been available there have been many changes in temperature trends. It seems imprudent to attempt to avoid or hide the obvious general positive trend which itself might or might not represent the trend over a longer period.
The proponents of the 'no warming' assertion express doubts about the probity of the terrestrial temperature records such as the HADCRUT3 from the UK Met Office. Figure 2 shows a comparison of the UAH and HADCRUT3 records since 1979.
Figure 2 UAH and HADCRUT3 records compared
The two sets of data correlate very closely with a correlation coefficient of 0.82. The equation of the best straight line drawn through the UAH data is: ΔT = 0.00115x - 0.228, where ΔT is the temperature anomaly and x represents the number of months since January 1979. The positive trend equates to 0.00115 × 12 = 0.014°C per annum or 1.4°C per century for a linear extrapolation.
The best line drawn through the HADCRUT3 data is: ΔT = 0.0013x + 0.001, and the positive trend is 0.00126 × 12 = 0.015°C per annum or 1.5°C per century for a linear extrapolation. The trends of the two sets of data are positive and almost identical and the difference between the intercepts is due to the different bases that they have, HADCRUT3 being 0.229°C higher than UAH.
A proper conclusion is that both sets of data are measurements of the same phenomenon; that of the rising temperature with time over the relatively small range of 32 years and that their trends might or might not be representative of future changes. The attribution of the positive temperature trends to the rising atmospheric concentration of CO2 is dealt with extensively on other pages of this site and in the scientific literature.
Is the world cooling down?
Much is being made of the observed cooling of the globe since 1998 with arguments such as: 'If the world is cooling and the atmospheric concentration of CO2 is rising, there must be something wrong with the physics'...
A statistical study by Liebmann, Dole, Jones, Bladé, and Allured in Bull. Am. Met. Soc., 91, 1485 (2010) is entitled 'Influence of Choice of Time Period on Global Surface Temperature Trend Estimates'. They take the HADCRUT3 and CRUTEM3 data sets for mean global and mean global land temperatures respectively from 1850 to 2009 and compute all the temperature trends from 2-year periods, 3-year periods, etc., to the overall 159-year period representing the maximum period in the data. The <159-year periods were computed for every possible starting year. No stone was unturned in the process!
A clear upward trend is evident in these series, both for the entire record and for many shorter segments. Large interannual-to-decadal variability is also apparent. These substantial short-term variations can lead to marked differences in trend estimates for time intervals whose starting and ending dates differ by only a few years. It is evident that time segments of a few decades or shorter can exhibit either warming or cooling trends, while trends for longer segments are mostly positive, though quite weak compared to those present in shorter segments.
Clearly, positive changes dominate the longer segments. The largest changes are positive and occur for segments longer than 30 years ending in recent years. The maximum change (0.82°C) is observed for the 108-yr period ending in 2009. There is also a secondary peak in warming that took place from the early twentieth century to the mid-1940s, culminating in a 0.62°C rise for the 39-yr period ending in 1945. For the sub-period 1945-2009 every trend longer than 22 years is positive, while for the entire record all segments longer than 82 years exhibit a positive trend. The overall warming trend, however, is interrupted by brief periods of cooling. From 1900 onwards, these cooling periods have not lasted more than 19 years, with two exceptions. The most pronounced episode of extended cooling took place following the warm interlude of 1937-45. When these warm years occur near the beginning of a segment, their influence on the trends extends all the way into the 1980s. There is some question, however, as to the accuracy of sea surface temperatures in the middle 1940s. A second, shorter period of cooling took place after the high temperatures of the late 1950s and early 1960s. The recent cooling that has been the subject of much popular media attention is presently of 9 years duration and amounts to a change of −0.07°C (the 4-8-yr changes ending in 2009 are also all negative). In total, there are 98 positive and 54 negative 9-yr segments. The segment ending in 2009 is only the 44th most negative of the record, well within the range of historical variability. Short-term trends of such magnitude (of either sign), therefore, are far from unusual.
Every possible trend (longer than two years) and its associated linear temperature change are calculated for the available record. Changes for segments longer than 82 years have all been positive. Within the constraints of the statistical significance tests, the positive changes of long duration (several decades and longer) ending in recent years are determined to be extremely unlikely to have occurred by chance. A secondary transient peak reflecting warming over a roughly 40-yr period ending in the 1940s is also unlikely to have occurred by chance. Since 1945, all periods longer than 22 years indicate warming, although only those segments ending recently stand out significantly from the noise. The land surface exhibits larger long-term changes than the entire globe, but the transient warming ending in the 1940s is less pronounced. In contrast, changes shorter than a few decades can be either positive or negative. The recent cooling trend is evident in the global record beginning in 2001. Such changes, however, are not statistically significant and are in fact relatively common in the historical record.
The paper makes it quite clear that there has been significant warming over the 1850-2009 period and opposite conclusions based upon any one set of ten consecutive years of data have a 35% chance of being wrong as representative of the general trend. In any case, the accepted definition of climate as the mean of thirty years of weather indicates the folly of attempting to come to conclusions about temperature trends using less than thirty consecutive years of data.
A New Study shows that warming is continuing
A new study published online today in the journal Science finds that the rate of global warming during the last 15 years has been as fast as or faster than that seen during the latter half of the 20th Century. The study refutes the notion that there has been a slowdown or "hiatus" in the rate of global warming in recent years.
The study is the work of a team of scientists from the National Oceanic and Atmospheric Administration's (NOAA) National Centers for Environmental Information (NCEI) using the latest global surface temperature data.
"Adding in the last two years of global surface temperature data and other improvements in the quality of the observed record provide evidence that contradict the notion of a hiatus in recent global warming trends," said Thomas R. Karl, L.H.D., Director, NOAA's National Centers for Environmental Information. "Our new analysis suggests that the apparent hiatus may have been largely the result of limitations in past datasets, and that the rate of warming over the first 15 years of this century has, in fact, been as fast or faster than that seen over the last half of the 20th century."
The apparent observed slowing or decrease in the upward rate of global surface temperature warming has been nicknamed the "hiatus." The Intergovernmental Panel on Climate Change's (IPCC) Fifth Assessment Report, released in stages between September 2013 and November 2014, concluded that the upward global surface temperature trend from 1998-2012 was markedly lower than the trend from 1951-2012.
Since the release of the IPCC report, NOAA scientists have made significant improvements in the calculation of trends and now use a global surface temperature record that includes the most recent two years of data, 2013 and 2014--the hottest year on record. The calculations also use improved versions of both sea surface temperature and land surface air temperature datasets. One of the most substantial improvements is a correction that accounts for the difference in data collected from buoys and ship-based data.
Prior to the mid-1970s, ships were the predominant way to measure sea surface temperatures, and since then buoys have been used in increasing numbers. Compared to ships, buoys provide measurements of significantly greater accuracy. "In regards to sea surface temperature, scientists have shown that across the board, data collected from buoys are cooler than ship-based data," said Dr. Thomas C. Peterson, principal scientist at NOAA's National Centers for Environmental Information and one of the study's authors. "In order to accurately compare ship measurements and buoy measurements over the long-term, they need to be compatible. Scientists have developed a method to correct the difference between ship and buoy measurements, and we are using this in our trend analysis."
In addition, more detailed information has been obtained regarding each ship's observation method. This information was also used to provide improved corrections for changes in the mix of observing methods.
New analyses with these data demonstrate that incomplete spatial coverage also led to underestimates of the true global temperature change previously reported in the 2013 IPCC report. The integration of dozens of data sets has improved spatial coverage over many areas, including the Arctic, where temperatures have been rapidly increasing in recent decades. For example, the release of the International Surface Temperature Initiative databank, integrated with NOAA's Global Historical Climatology Network-Daily dataset and forty additional historical data sources, has more than doubled the number of weather stations available for analysis.
Lastly, the incorporation of additional years of data, 2013 and 2014, with 2014 being the warmest year on record, has had a notable impact on the temperature assessment. As stated by the IPCC, the "hiatus" period 1998-2012 is short and began with an unusually warm El Niño year. However, over the full period of record, from 1880 to present, the newly calculated warming trend is not substantially different than reported previously (0.68°C / Century (new) vs 0.65°C / Century (old)), reinforcing that the new corrections mainly have in impact in recent decades.
JB: A more judicious line would have started in 1976, the reasonable end year for the cooling since 1945.