I'm not sure how you can make that claim with only 29 years of data without making some pretty big assumptions about the underlying distribution.
We have ways to determine the past climate without having access to direct measurements.
We have ways to determine the past climate without having access to direct measurements.