The latest suite of climate models collectively projects that the average temperature of the global atmosphere should be increasing at the rate of about 0.28–0.29 degrees Celsius per decade. But how do these model projections compare to actual temperature data? It depends.
Let's first take a look at research using surface thermometer data assembled from weather stations, ocean-going ships, and buoys. The Berkeley Earth team reports that since 1980, the global average temperature is increasing at the rate of 0.19 degrees Celsius per decade. The National Oceanic and Atmospheric Administration (NOAA) finds that the global average temperature has been increasing at the rate of 0.18 degrees Celsius per decade since 1981. NASA's GISTEMP data set reports an increase of 0.19 degrees Celsius per decade. The U.K.'s Hadley Centre finds the increase is about 0.20 degrees Celsius per decade.
The European Centre for Medium-Range Weather Forecasts reports the global average temperature trend generated by its fifth-generation atmospheric reanalysis (ERA5). Reanalysis is a blend of observations with past short-range weather forecasts rerun with modern weather forecasting models. From 1979 on, the ERA5 calculates that the global average temperature has been increasing at a rate of 0.19 degrees Celsius per decade. The Japan Meteorological Agency's JRA-55 reanalysis finds the per-decade rate of increase is 0.18 degrees Celsius.
Climate researchers also have access to temperature data sets derived from satellite measurements that essentially measure temperature trends in the whole atmosphere (troposphere) beginning in 1978. The first satellite data set was devised by University of Alabama-Huntsville (UAH) climate researchers John Christy and Roy Spencer. According to UAH measurements, the rate of global average temperature increase is running at 0.13 degrees Celsius per decade.
Researchers don't just read numbers off satellite feeds to discover temperature trends. They must take into account the orbital decay of satellites, the deterioration of instruments, and changes related to replacing satellites over time. Another team of researchers at Remote Sensing Systems has parsed the satellite data and derived a tropospheric temperature trend of 0.18 degrees Celsius per decade. Clearly, this more closely matches the surface thermometer trends.
In March, another team associated with NOAA's Center for Satellite Applications and Research (STAR) reported in the Journal of Geophysical Research: Atmospheres its analysis of the satellite temperature data. Earlier, the STAR researchers had calculated that the temperature trend for the total troposphere (TTT) was about 0.16 degrees Celsius per decade. After making further adjustments, the STAR team in March lowered the trend to a bit over 0.14 degrees Celsius per decade.
"The total TTT trend found in this study was only one-half of the climate model simulations," the STAR researchers note. "Possible reasons for the observation-model differences in trends may include climate model biases in responding to external forcings, deficiencies in the post-millennium external forcings used in model simulations, phase mismatch in natural internal climate variability, and possible residual errors in satellite data sets." Translation: The models simply run too hot, the historical inputs like volcanic aerosols and ozone to the models may be wrong, a temporary natural cooling trend could be masking warming, and adjustments to the satellite data may be wrong.
The STAR researchers tellingly add that their findings are "consistent with conclusions in McKitrick and Christy (2020) for a slightly shorter period (1979–2014)." In that 2020 study, environmental economist Ross McKitrick and Christy compared the outputs of the latest suite of climate models to satellite, weather balloon, and reanalysis products. They found that every one of the 38 new generation "climate models exhibits an upward bias in the entire global troposphere as well as in the tropics." The models are predicting much more warming than appears to be occurring. Again, they are running too hot.
The new STAR study researchers do additionally observe, "A striking feature is that trends during the latest half period (around 0.21–0.22 K/decade) nearly doubled the trends during the first half period (around 0.10–0.12 K/decade) for the global and global ocean means. These large differences in TTT trends between the first and second half periods suggest that the tropospheric warming is accelerating." It is worth noting that this accelerated trend is still about a third lower than the average of the model projections.
However, McKitrick in a preliminary analysis over at Climate Etc. finds, "the new NOAA data do not support a claim that warming in the troposphere has undergone a statistically-significant change in trend."
Given that climate science is continually evolving, it's a good idea to heed University of Colorado climate policy researcher Roger Pielke Jr.'s admonition to "be careful celebrating the results of any one study too much, because science moves ahead and there is no guarantee that any single paper stands the test of time."
In his comparison of new STAR data with other temperature data sets, NASA climate modeler Gavin Schmidt gamely points out, "The upward trends differ slightly for sure, but they are all recognizably describing the same climate change." But, in fact, all of the surface and satellite temperature trends are considerably lower than the average of the projections made by the most recent set of climate models.
Average global temperature has increased by about 1.1 degrees Celsius since the late 19th century. If the rate of warming is not in fact accelerating, rough extrapolations of the lowest and highest rates of warming derived from the observational records suggest that unabated global warming would further boost average temperatures between 1 and 1.6 degrees Celsius by the end of this century. Such an increase is in line with recent research that finds that the average global temperature is likely to rise by 2100 to about 2.2 degrees Celsius above the 19th century baseline. That's not nothing, but such an increase is unlikely to be catastrophic for future generations.
Start your day with Reason. Get a daily brief of the most important stories and trends every weekday morning when you subscribe to Reason Roundup.