Materials of geological origin will have absorbed considerable quantities of radiation since their formation, so any human-caused exposure to heat or light will reset the luminescence clock considerably more recently than that since only the energy stored since the event will be recorded.
The possibility of making use of TL stored in a mineral or pottery sample was first proposed by chemist Farrington Daniels in the 1950s. Applications and limitations of thermoluminescence to date quaternary sediments.
We conclude that while TL dating has the potential to solve many sedimentary problems, more fundamental research needs to be carried out before such dates should be accepted.
A set of criteria for acceptable dates is proposed.
The equations used by scholars to determine the date when the last exposure happened are typically: Where De is the laboratory beta dose that induces the same luminescence intensity in the sample emitted by the natural sample, and DT is the annual dose rate comprised of several components of radiation that arise in the decay of natural radioactive elements.
See Liritzis et al.'s excellent 2013 book on Luminescence Dating for more information on these processes.