INTRODUCTION
Athol Rafter and Gordon Fergusson started atmospheric 14CO2 measurements near Wellington, New Zealand, in late 1954 (Rafter Reference Rafter1955a). Initially, the goal was to improve 14C as a dating tool by investigating the natural 14C levels in the environment, including the atmosphere, surface ocean, and biosphere (Rafter Reference Rafter1955a). In a 1965 after-dinner speech at the Sixth International Radiocarbon and Tritium Conference in Pullman, Washington, Rafter recounted his introduction to the world of radiocarbon dating: “A few days later I was walking home quietly through the grounds of Parliament Buildings when coming in the opposite direction was the Head of our Department, a Mr Callaghan, who stopped me with the statement, ‘Rafter, I have just come from a meeting with geologists who tell me there is a method of dating by means of carbon that should be able to tell the age of our volcanic ash showers. Would you see if you could develop this method and stop the geologists arguing?’ I said a confused goodnight and continued on my way home somewhat more puzzled than usual” (Rafter Reference Rafter1965).
It wasn’t long before Rafter and compatriots Gordon Fergusson and George Page had an operational 14C lab, making their first 14C measurements in 1951 (Fergusson and Rafter Reference Fergusson and Rafter1953). These first measurements constrained the age of the Taupo supervolcano eruption “volcanic ash shower” and its ubiquitous tephra layer throughout New Zealand and the South Pacific, to about 170 AD. Unfortunately, these hard-won results did not stop the geologists arguing. The age was further constrained to 230 AD (Sparks et al. Reference Sparks, Melhuish, McKee, Ogden, Palmer and Molloy1995), refined to 232 ± 5 AD (Hogg et al. Reference Hogg, Lowe, Palmer, Boswijk and Ramsey2011), and disputed again more recently (Holdaway et al. Reference Holdaway, Duffy and Kennedy2018; Hogg et al. Reference Hogg, Wilson, Lowe, Turney, White, Lorrey, Manning, Palmer, Bury and Brown2019). Interestingly, when the 1951 measurements are calibrated, the result comes out remarkably close to the later results at 245 AD, albeit with a larger uncertainty (Sparks Reference Sparks2004).
Nonetheless, the arguing geologists no doubt inspired the early researchers in their efforts to achieve the highest quality and precision. Libby and his colleagues had achieved around 2% precision with their solid carbon measurement system (Anderson et al. Reference Anderson, Libby, Weinhouse, Reid, Kishenaum and Grosse1947; Libby et al. Reference Libby, Anderson and Arnold1949), but Rafter and Fergusson were stymied in their early attempts by the state of the local roads (Rafter Reference Rafter1953). Rafter would prepare solid samples at his laboratory in Wellington city, but the process of transporting the samples the 15 km or so to Fergusson’s counters in Lower Hutt (near the current location of the Rafter Radiocarbon Laboratory) would disappointingly result the transfer of the carbon carefully adhered to the walls of a copper cylinder to a pile of unusable rubble in the bottom of the container. Thus a better method was needed, and Rafter and Fergusson were the first to develop a CO2 gas counting system, making their first measurements in 1951 (Rafter Reference Rafter1955b), although European researchers working on a similar system published their work earlier (see in this issue Grootes and van der Plicht Reference Grootes and van der Plicht2021).
The CO2 counting technique was not only easier and less dangerous, but the measurement precision was markedly improved over the earlier designs, achieving around 0.5% precision (de Vries and Barendsen Reference de Vries and Barendsen1953; Rafter Reference Rafter1955a) vs. the 2% obtained by Libby and others (Anderson et al. Reference Anderson, Libby, Weinhouse, Reid, Kishenaum and Grosse1947). It was this improved precision that allowed detection of the beginning of the 14C bomb spike from first three atmospheric 14C measurements in 1954 and 1955 (Table 1, Figure 1). Yet “the CO2 method [was] a most beautiful method and like all things beautiful has a most cantankerous side, the presence of electronegative impurities that drive technicians to despair, scientists to drink and harmony to discord” (Rafter Reference Rafter1965).
EARLY ATMOSPHERIC 14CO2 RESULTS
The very first measurements at the New Zealand laboratory quickly led to a recognition that atmospheric 14CO2 measurements would be invaluable in understanding the carbon cycle itself. Svante Arrhenius had worked out that fossil fuel emissions could double the amount of CO2 in the atmosphere but expected it to take a few thousand years (Arrhenius Reference Arrhenius1896). Rafter’s early measurements (Rafter and Fergusson Reference Rafter and Fergusson1958), along with tree-ring reconstruction of 1940s and 1950s atmospheric 14CO2 (Suess Reference Suess1955), demonstrated a decrease in 14C in the atmosphere that was hypothesized to be due to the absence of 14C in the carbon from fossil fuel burning that had been added to the atmosphere (Revelle Reference Revelle1986), and that it was happening about 30 times faster than Arrhenius had expected (Suess Reference Suess1955). Both Rafter and Suess recognized there was a need to measure atmospheric CO2 mole fractions more directly, leading to the initiation of the iconic CO2 mole fraction record at Mauna Loa, Hawai’i in 1958 (Keeling Reference Keeling1960). These discoveries by Arrhenius, Rafter, and Suess were the first two major science discoveries about climate change, and the fact that global changes can happen faster than expected in some early science papers has been rediscovered many times since.
Soon after, Rafter and Fergusson’s research led to another discovery, that the surface ocean appears about 400 years older than the overlying atmosphere (Rafter Reference Rafter1955a; Craig Reference Craig1957). This key research set the stage for the marine radiocarbon calibration research field (Heaton et al. Reference Heaton, Köhler, Butzin, Bard, Reimer, Austin, Bronk Ramsey, Grootes, Hughen and Kromer2020), and was the foundation for the use of 14C in understanding the large-scale ocean circulation (Broecker et al. Reference Broecker, Peng, Ostlund and Stuiver1985).
Rafter was surprised by the first few atmospheric 14CO2 measurement results, which showed a significant upward trend: “Whether this is due to an experimental error in the method of collection or to some other factor will be discussed when a greater number of samples have been studied” (Rafter Reference Rafter1955a). In fact, these were the first observations of the atmospheric 14CO2 increase due to atmospheric nuclear weapons testing. Apparently Rafter already suspected the cause in 1955, in 1957 saying that “sampling was continued to test whether or not there was a seasonal variation or a possible 14C enrichment of the atmosphere from atomic explosions” (Rafter and Fergusson Reference Rafter and Fergusson1957a). In an unpublished paper delivered at a conference in Dunedin, New Zealand in 1957, Rafter stated, “I would like to tell you the possibilities of two gigantic tracer experiments that we are witnessing in the world today and how if we are quick enough, we will be able to solve some interesting problems in mass movements of interest to oceanography and meteorology.” He foresaw the use of careful measurements of both the Suess and the bomb effects as valuable tools in geophysical research. An unverified report claims that during an international conference in the mid-1950s, Athol Rafter prudently withdrew a presentation that would have demonstrated the link with nuclear weapons testing after a late-night visit from government agents. Not daunted for long, Rafter and Fergusson published a paper entitled “The atom bomb effect” in the New Zealand Journal of Science and Technology in 1957, and then shared the results more widely in Science later that same year (Rafter and Fergusson Reference Rafter and Fergusson1957a; Rafter and Fergusson Reference Rafter and Fergusson1957b). Rafter claimed that should secret atmospheric nuclear testing be carried out in the Pacific, his 14CO2 measurements meant that New Zealand would know about it within days (Priestley Reference Priestley2012).
A 1965 letter from Hans Suess (at the University of California, San Diego) to Athol Rafter reads, “it is embarrassing for many of us that the rise in the carbon-14 concentration in the atmosphere due to artificial sources was first discovered and quantitatively measured by your laboratory for the southern hemisphere.” This dose of healthy competition was perhaps the spark for Northern Hemisphere researchers to begin atmospheric 14CO2 observations later in the 1950s and early 1960s. Certainly, the measurements from both hemispheres have provided a treasure trove of data that continues to be applied across multiple disciplines.
THE LEGACY OF EARLY ATMOSPHERIC 14C MEASUREMENTS
Once the first atmospheric 14C measurements had demonstrated an increase in 14CO2 content, attributed to atmospheric nuclear weapons testing, several laboratories became interested in this phenomenon and took up tropospheric 14CO2 measurements at numerous locations around the world (Nydal and Lövseth Reference Nydal and Lövseth1983; Berger et al. Reference Berger, Jackson, Michael and Suess1987; Levin et al. Reference Levin, Kromer, Schoch-Fischer, Bruns, Munnich, Berdau, Vogel and Munnich1985; Manning et al. Reference Manning, Lowe, Melhuish, Sparks, Wallace, Brenninkmeijer and McGill1990; Nydal and Gislefoss Reference Nydal and Gislefoss1996) as well as in the stratosphere (Telegadas et al. Reference Telegadas, Gray, Sowl and Ashenfelter1972). More recently, detonation locations and magnitudes have been matched up with the 14CO2 observations to estimate the total 14C production from nuclear weapons (Hesshaimer and Levin Reference Hesshaimer and Levin2000; Naegler and Levin Reference Naegler and Levin2006), a key parameter in establishing global radiocarbon budgets.
Once the threat of catastrophic 14C production had passed, many atmospheric 14CO2 measurement records were halted, likely due more to funding constraints than a lack of interest on the part of researchers. Rafter and his team continued the Wellington record (Manning et al. Reference Manning, Lowe, Melhuish, Sparks, Wallace, Brenninkmeijer and McGill1990), moving the sampling site from the original Makara location on New Zealand’s western coast near Wellington. The new site at Baring Head, on the south coast near Wellington, proved to be consistent with Makara for 14C content, but a better choice for CO2 mole fraction measurements, as it is much less influenced by the local biosphere (Lowe Reference Lowe1974; Lowe et al. Reference Lowe, Guenther and Keeling1979). The Wellington record continues to this day, its 67 years of measurements representing the longest direct atmospheric trace gas or isotope record anywhere in the world. In the 1990s, Rafter’s original gas counting method was phased out and replaced with AMS. The New Zealand laboratory contributed substantially to the development of AMS, particularly for atmospheric applications and development of the hydrogen graphitization technique (Lowe Reference Lowe1984; Lowe et al. Reference Lowe, Brenninkmeijer, Manning, Sparks and Wallace1988). The atmospheric record continued to use the sodium hydroxide absorption method of sample collection to ensure continuity of the record and these measurements have been supplemented with whole air flask samples since 2012. Fortuitously, Rafter’s successors including Graeme Lyon and Gordon Brailsford had collected and archived CO2 from whole air flasks in the late 1980s and early 1990s, and these were able to replace anomalous outlier data for that period. Less fortunately, the early AMS measurements from 1995 to 2005 were made with incomplete correction for isotopic fractionation during sample preparation and measurement, and subsequent analysis has demonstrated that these data are biased high. Tree-ring measurements from Baring Head and nearby in the 2010s demonstrated that the early gas counting measurements from 1954 through to the late 1980s did not display any detectable biases or problems, a real testament to the careful efforts of the early researchers at a time when there were few options to validate or check their data (Turnbull et al. Reference Turnbull, Mikaloff Fletcher, Ansell, Brailsford, Moss, Norris and Steinkamp2017).
The other notable uninterrupted time series of 14CO2 are from the European Alps, initially at the Austrian high altitude site of Vermunt (Levin et al. Reference Levin, Kromer, Schoch-Fischer, Bruns, Munnich, Berdau, Vogel and Munnich1985). Like the Wellington record, the sampling site was moved in the 1980s, in this case to Jungfraujoch in Switzerland, an even higher altitude site more favorable for CO2 and other trace gas measurements (see in this issue Levin et al. Reference Levin, Hammer, Kromer, Preunkert, Weller and Worthy2021).
Together these Southern and Northern Hemisphere records have provided a strong constraint on interhemispheric mixing times (Kjellström et al. Reference Kjellström, Feichter and Hoffman2000; Land et al. Reference Land, Feichter and Sausen2002). These long records are key to constraining the global radiocarbon budget and elucidating carbon exchange processes (Randerson et al. Reference Randerson, Enting, Schuur, Caldeira and Fung2002; Naegler and Levin Reference Naegler and Levin2006; Levin et al. Reference Levin, Naegler, Kromer, Diehl, Francey, Gomez-Pelaez, Steele, Wagenbach, Weller and Worthy2010). They provide the backdrop for determining the rate of uptake of anthropogenic carbon into the oceans (Broecker et al. Reference Broecker, Peng, Ostlund and Stuiver1985; Hesshaimer et al. Reference Hesshaimer, Heimann and Levin1994; Caldeira et al. Reference Caldeira, Rau and Duffy1998; Key Reference Key2004; Peacock Reference Peacock2004; Sweeney et al. Reference Sweeney, Gloor, Jacobson, Key, McKinley, Sarmiento and Wanninkhof2007), as well as in understanding the gas exchange rate with the oceans (Krakauer et al. Reference Krakauer, Randerson, Primeau, Gruber and Menemenlis2006). A renewed interest in atmospheric 14CO2 observations since the turn of the century has allowed further investigation into Southern Ocean carbon exchange, demonstrating that upwelled deep waters drive an atmospheric latitudinal gradient in 14CO2 (Levin et al. Reference Levin, Kromer, Wagenbach and Münnich1987, Reference Levin, Naegler, Kromer, Diehl, Francey, Gomez-Pelaez, Steele, Wagenbach, Weller and Worthy2010; Graven et al. Reference Graven, Gruber, Key, Khatiwala and Giraud2012a, Reference Graven, Guilderson and Keeling2012b).
The atmospheric measurements are also leveraged in soil radiocarbon studies, for which the 14C bomb spike provides a vital control on turnover times (Trumbore Reference Trumbore2000; Eglinton et al. Reference Eglinton, Galy, Hemingway, Feng, Bao, Blattmann, Dickens, Gies, Giosan and Haghipour2021). Perhaps more familiar to the radiocarbon dating community, the direct atmospheric 14CO2 records form the backbone of the “calibomb” dataset, used to calibrate post-bomb 14C measurements since 1950 (Hua et al. Reference Hua, Barbetti and Rakowski2013, Reference Hua, Turnbull, Santos, Rakowski, Ancapichun, de Pol-Holz, Hammer, Lehman, Levin, Miller, Palmer and Turney2021). The measurements of the rapid changes in atmospheric 14CO2 content since 1950 mean that samples can be aged quite precisely, often to within a single year. A wealth of applications include authentication of art objects; testing confiscated elephant ivory for violations of the CITES agreement (Uno et al. Reference Uno, Quade, Fisher, Wittemyer, Douglas-Hamilton, Andanje, Omondi, Litoroh and Cerling2013; Cerling et al. Reference Cerling, Barnette, Chesson, Douglas-Hamilton, Gobush, Uno, Wasser and Xu2016) and other wildlife forensic studies (Uno et al. Reference Uno, Quade, Fisher, Wittemyer, Douglas-Hamilton, Andanje, Omondi, Litoroh and Cerling2013); authentication of bioplastic source (Telloli et al. Reference Telloli, Rizzo, Canducci and Bartolomei2019; Rogers et al. Reference Rogers, Turnbull, Dahl, Phillips, Bridson, Raymond, Liu, Yuan and Hill2021); and even as a tool for determining the age of human remains for forensic investigations.
While Rafter and his compatriots were able to observe the changes in 14CO2 in the clean Southern Hemisphere air near Wellington, they also noted decreases in 14CO2 content. This was the imprint of 14C-free fossil fuel CO2 introduced into the atmosphere, now coined the “Suess Effect” (Suess Reference Suess1955). Several early papers documented the Suess Effect in pre-bomb tree-ring samples (Lerman et al. Reference Lerman, Mook and Vogel1970; Tans et al. Reference Tans, De Jong and Mook1979; Stuiver and Braziunas Reference Stuiver and Braziunas1998), but the enormous perturbation of the bomb spike masked the Suess Effect for much of the latter half of the 20th century. The renewed enthusiasm for atmospheric 14CO2 measurements in the last 20 years is primarily focused on 14CO2, widely recognized as the “gold standard” tracer for fossil fuel CO2 emissions. The clean air measurements such as those at Wellington and Jungfraujoch provide a background constraint when determining the recently added fossil fuel CO2 component. This application has become widespread, used to determine fossil fuel CO2 emissions from individual point sources (Turnbull et al. Reference Turnbull, Keller, Norris and Wiltshire2016), cities (Djuricin et al. Reference Djuricin, Pataki and Xu2010; Ding et al. Reference Ding, Shen, Yi, Wang, Ding, Fu and Liu2013; Newman et al. Reference Newman, Xu, Gurney, Hsu, Li, Jiang, Keeling, Feng and Keefe2016; Niu et al. Reference Niu, Zhou, Cheng, Wu, Lu, Xiong, Du and Fu2016; Turnbull et al. Reference Turnbull, Karion, Fischer, Faloona, Guilderson, Lehman, Miller, Miller, Montzka and Sherwood2011, Reference Turnbull, Karion, Davis, Lauvaux, Miles, Richardson, Sweeney, McKain, Lehman and Gurney2018), to nations and regions (Levin et al. Reference Levin, Kromer, Schmidt and Sartorius2003; Levin and Kromer Reference Levin and Kromer2004; Hsueh et al. Reference Hsueh, Krakauer, Randerson, Xu, Trumbore and Southon2007; Palstra et al. Reference Palstra, Karstens, Streurman and Meijer2008; Riley et al. Reference Riley, Hsueh, Randerson, Fischer, Hatch, Pataki, Wang and Goulden2008; Van Der Laan et al. Reference Van Der Laan, Karstens, Neubert, Van Der Laan-Luijkx and Meijer2010; Cui et al. Reference Cui, Newman, Xu, Andrews, Miller, Lehman, Jeong, Zhang, Priest and Campos-Pineda2019; Basu et al. Reference Basu, Lehman, Miller, Andrews, Sweeney, Gurney, Xu, Southon and Tans2020; Lee et al. Reference Lee, Dlugokencky, Turnbull, Lee, Lehman, Miller, Pétron, Lim, Lee and Lee2020).
This ongoing addition of 14C-free fossil-fuel CO2 is predicted, if emissions continue apace, to result in an apparent atmospheric radiocarbon age of around 2500 years in 2100 (Graven Reference Graven2015). While not the most dire consequence of human perturbations to the atmosphere, it will certainly be an inconvenience to the radiocarbon dating community.
CONCLUSIONS
The sheer breadth of ongoing applications for the atmospheric 14CO2 measurements can be seen in the citation rate. The Wellington 14CO2 record started by Athol Rafter has been cited directly (Rafter and Fergusson Reference Rafter and Fergusson1957b; Manning et al. Reference Manning, Lowe, Melhuish, Sparks, Wallace, Brenninkmeijer and McGill1990; Currie et al. Reference Currie, Brailsford, Nichol, Gomez, Sparks, Lassey and Riedel2011; Turnbull et al. Reference Turnbull, Mikaloff Fletcher, Ansell, Brailsford, Moss, Norris and Steinkamp2017) and indirectly through the calibomb dataset (Hua and Barbetti Reference Hua and Barbetti2004; Hua et al. Reference Hua, Barbetti and Rakowski2013) more than 1000 times. In recognition of his huge influence on the development of atmospheric 14CO2 and radiocarbon measurements in general, the Wellington laboratory was renamed the Rafter Radiocarbon Laboratory in 1993 in celebration of Athol Rafter’s 80th birthday (Sparks Reference Sparks2004).
With the new technologies introduced in the last few years, and the prospect of in situ 14C measurements on the horizon, atmospheric 14CO2 measurements can be expected to expand even further in the coming decades. Athol Rafter and his fellow researchers who began these measurements might be surprised at how widely used their ideas and data have become. But perhaps not, in Rafter’s own words, “if these 14C increases in the main reservoirs of the carbon cycle can be adequately studied in both hemispheres, it would be possible to evaluate exchange constants across the stratospheric-tropospheric, tropospheric-surface ocean water, surface ocean – deep ocean water, and interhemispheric differences” (Rafter and Fergusson Reference Rafter and Fergusson1957a).
ACKNOWLEDGMENTS
In addition to the lead scientists who developed the atmospheric 14CO2 record in New Zealand, many technicians and assistants made vital contributions: Max Burr, Charlie McGill, Dave Currie, Gordon Brailsford, Hank Janssen. The GNS librarians, particularly Maggie Dyer and Pauline Muir, unearthed the early papers, reports and letters cited in this report. Dawn Chambers rescued many of the original paper records from a skip bin when the original Shed 2 lab was dismantled, and painstakingly documented countless records from the early period. Stephen Suess kindly provided permission to publish excerpts from his father’s letters.