They aren't guesses. Those charts are reflecting oxygen isotope ratios (which are measured directly), among countless others (which are covered in Hasen et al)

Oxygen isotope ratios are more or less ^16O and ^18O. These isotopes are distributed differently based on temperature and precipitation patterns, which can be recorded in ice cores, ocean sediments, lake sediments, and even in the shells of marine organisms. ^16O is more often seen in 'warmer' events, vs ^18O