“It was surprising, really surprising to me,” said Benjamin Santer, a climate scientist at the Woods Hole Oceanographic Institution, “that we could have identified with high confidence a human-caused stratospheric cooling signal within 25 years of the start of monitoring, if we had back then in 1860 the measuring capability that we have today.”
That capability now with satellite-borne microwave radiometers, weather balloons, and climate models has been employed retroactively. In Proceedings of the National Academy of Sciences, researchers wedded historical records, physical theory, and supercomputer simulations to recreate historical changes in atmospheric temperatures at the start of the industrial era. The scientists wanted to find out when the “fingerprint” of human activity first might have been detected in the vertical profile of the atmosphere.
Fingerprinting is a pattern recognition technique comparing observed temperature changes to model simulations with and without anthropogenic forcings. In the stratosphere 15 to 50 kilometers above Earth the signal manifests as severe cooling, a radiative impact of increasing carbon dioxide levels. The stratosphere is less subject to short-term weather fluctuations than is the lower atmosphere, making it the best layer to detect early on.
Employing nine climate models and reconstructions of past greenhouse gas levels from ice cores and direct measurements, the team concluded that if current monitoring networks existed in 1860, the human fingerprint would have been detectable by around 1885. Even with hemispheric mid-latitude coverage only, detection would have been feasible by 1894. Atmospheric CO₂ had increased only about 10 parts per million in four decades at that time yet the signal was already robust enough to be detectable above natural variability.
Current satellite measurements, now available from 1979 for the lower stratosphere and from 1986 for the mid-to-upper stratosphere, confirm the predicted pattern: tropospheric warming and cooling with altitude above, with cooling augmented with increasing altitude. Adding mid-to-upper stratospheric data improves detectability by five compared to analyses cut off at 25 kilometers altitude. This is due to the fact that both the magnitude of the cooling roughly 1.8 to 2.2°C globally since the mid-1980s and natural process “noise” at such altitudes are low.
Satellite temperature measurements must be calibrated with extreme care. Drifting orbits, aging instruments, and new satellites cross-calibrated against their older counterparts. Yet when corrected, they provide a nearly uninterrupted global view immensely more valuable than weather balloons’ sporadic and local sampling. When compared to climate model simulations with and without the effects of human activities, scientists are able to calculate the likelihood that patterns seen are the result of natural variability. In this case, this likelihood is infinitesimally small.
The study also points out the contribution of the development of climate modeling to the reconstruction of the past. By simulating the climate system response to anthropogenic and natural forcing agents volcanic activity and solar cycles, for example models can recreate the interplay among signals that would have been present in the 19th century. For example, the 1883 Krakatoa volcanic eruption briefly dominated the CO₂-induced stratospheric cooling signal, delaying detectability by a few years in the simulations.
Since that initial measurable moment, CO₂ in the atmosphere has increased by approximately 140 parts per million. The early emergence of the fingerprint suggests that, if 19th-century scientists had possessed contemporary instruments, the direction of climate policy might have been fundamentally transformed. As University of Graz’s Andrea Steiner penned, “This confirms that temperature change signals of the atmosphere are effective not only for detection, but also as early indicators of the success of climate mitigation efforts.”
The finding arrives amid concerns of budget cuts to satellite programs at NASA and NOAA, whose loss Santer testifies would impair the ability to monitor current changes. The scientific take-home is unmistakable: the vertical temperature profile of the atmosphere its stair-step progression of warming troposphere and cooling stratosphere is even now one of the most rigorous tests for human forcing of climate, and the mid-to-upper stratosphere is its most revealing canvas.
