New GloSAT Dataset and the IPCC's +1.5°C Target

One of the main reasons why most of our GMST datasets go back only to 1850 is that we don't have enough SST measurements prior to 1850 to calculate GMST anomalies with a confidence interval small enough to be useful. However, ships did record surface air temperatures at least back into the 1780s. Recently the Global Surface Air Temperature project (GloSAT) digitized records of marine air temperatures taken on ships to form the GloSAT dataset, which contains near surface air temperatures back to 1781. Ed Hawkins, who worked on the project, has a good post describing the results of the project. The GloSAT database only goes to 2021, so I have included HadCRUT5 so that you can see the close agreement between the two. I find it interesting that the GloSAT dataset captures a little more variability in GMAT in the late 19th century, though trends are not that different between the two. The 95% confidence intervals for annual means are less than 0.3°C for most of the 18th century, except during the 1780s. The dip in GMAT between 1810 and 1816 was due to volcanic activity at the time.

GloSAT with HadCRUT5

The IPCC sets its targets as warming above preindustrial levels, and the industrial revolution began in around 1750. IPCC assessment reports have used 1850-1900 as a preindustrial baseline mostly because of data poverty prior to 1850, and proxy evidence indicates that GMST was not that much different between 1750 and 1850. Well, now we have a better idea by how much. I set this dataset to the 1790-1839 mean calculated that the 1790-1839 mean was 0.22°C cooler than the 1850-1900 mean.

GloSAT Shows +1.5°C Above Preindustrial Levels

I then calculated centered running 30-year means, and the trend for 1992-2021, which was 0.21°C/decade, and projected forward to 2025 to estimate how close we've come to the IPCC's +1.5°C Target when set to a1790-1839 baseline. I came up with almost exactly 1.5°C. I'm not sure we should say that we've reached that target yet though, since many of the estimates for for the consequences of crossing 1.5°C may be assuming an 1850-1900 baseline. I do think it's fair to say that we're knocking on the door of crossing the target, though.

References:

[1] Morice, C. P., Berry, D. I., Cornes, R. C., Cowtan, K., Cropper, T., Hawkins, E., Kennedy, J. J., Osborn, T. J., Rayner, N. A., Recinos Rivas, B., Schurer, A. P., Taylor, M., Teleti, P. R., Wallis, E. J., Winn, J., and Kent, E. C.: An observational record of global gridded near-surface air temperature change over land and ocean from 1781, Earth Syst. Sci. Data, 17, 7079–7100, https://doi.org/10.5194/essd-17-7079-2025, 2025.

[2] Morice, C.P.; Berry, D.I.; Cornes, R.C.; Cowtan, K.; Cropper, T.; Hawkins, E.; Kennedy, J.J.; Osborn, T.; Rayner, N.A.; Rivas, B.R.; Schurer, A.; Taylor, M.; Teleti, P.R.; Wallis, E.J.; Winn, J.P.; Kent, E.C. (2025): GloSATref.1.0.0.0: An observational record of global gridded near surface air temperature change over land and ocean from 1781. NERC EDS Centre for Environmental Data Analysis, 19 June 2025. doi:10.5285/a2519624a593402a83246bd359d098be. https://dx.doi.org/10.5285/a2519624a593402a83246bd359d098be

[3] Kennedy, John; Becker, Amani; Cowtan, Kathryn D; Hawkins, Ed; Hegerl, Gabi; Middleton, Stuart E.; et al. (2025). Illustration introducing the GloSAT Global Surface Air Temperature project. figshare. Figure. doi: 10.6084/m9.figshare.28001222