> You can't just keep doing this without suffering a loss in significant figures. I'm not sure I understand this. Are you talking about something like the numb

> You can't just keep doing this without suffering a loss in significant figures. I'm not sure I understand this. Are you talking about something like the number of significant figures in the naive arithmetic mean being that of the individual measurement with the fewest significant figures (which is more or less expected)? Or are you talking about some loss of significant figures beyond that? Papers describing models generally go into some amount of detail describing the precise averaging/interpolation/etc. methods and associated statistical analysis (e.g., [0, 1] for Berkeley Earth's model). If there are errors regarding the uncertainty analysis, that's where you're likely to find them. > You would literally have to have accuracy and precision information on each sensor and for each measurement to then be able to propagate all of this carefully. Furthermore, you really need to know the standard deviation in order to understand the uncertainty in the measurements. Is having that information that unbelievable, especially for scientific instruments? The field of metrology exists for a reason, and it's not like the need to understand the accuracy and precision of instruments is a recent concern. In addition, as I linked above and also in [2], uncertainty analyses for models do exist, and measurement uncertainties are absolutely taken into account. > What would the graph presented in the article look like with error bands? Don't know. A plot of the global annual land temperature anomaly over time [3] is present in the GISTEMP uncertainty analysis paper I linked. I'd expect that similar graphs could be found in relevant papers for other models. > so I'll stick to not being happy with the way things have been presented so far and hope to find something better in the future. For better or worse, presentation is a tradeoff. Simpler graphs are easier to digest for the average person, but lack the nuances someone like you might be looking for. More detailed graphs are (sometimes?) present in papers, but the average person isn't likely to come across those. Communication is hard. [0]: https://berkeleyearth.org/static/papers/Methods-GIGS-1-103.pdf https://berkeleyearth.org/static/papers/Methods-GIGS-1-103.p... [1]: https://berkeleyearth.org/static/pdf/methods-paper-supplement.pdf https://berkeleyearth.org/static/pdf/methods-paper-supplemen... [2]: https://www.st-andrews.ac.uk/~rjsw/papers/Brohan-etal-2006.pdf https://www.st-andrews.ac.uk/~rjsw/papers/Brohan-etal-2006.p... [3]: https://agupubs.onlinelibrary.wiley.com/cms/asset/a68bb281-c6fd-4b9e-9af0-54c20f9cd9f9/jgrd55486-fig-0006-m.jpg https://agupubs.onlinelibrary.wiley.com/cms/asset/a68bb281-c...