How accurate is Radiometric Dating?

[Anomalocaris]

It’s not the inconstancy of the rate of decay that puts unavoidable error bars on radiometrid dating.    It is the uncertainty surrounding just how much of the parent element there were at the moment of the event you are trying to date, whether certain quantity of daughter element had already been there at that moment,  and if there was some way we have not yet been able to think of or control for that allows some of the parent and daughter element to escape or be removed, thus screwing up the ratio of parent to daughter elements and thus the implied radiometrid age.

However, the possibility of such disturbances to parent-daughter element ration screwing up radiomatric dating can be reduced by cross check several different radioactive decay clocks, if these are available.     The chance that several radiometrid clocks are disturbed by the exact same amount so they all agree on the wrong age is virtually nil.

If several different radiometric clocks disagree in statisistically significant way, then you have some thinking, anssessment and experimentation to do to figure out which one(s) could have been disturbed.

It should also be pointed out that Rb-Sr dating is self-correcting in this way. You need to have several different crystals from the same rock, but you can derive the amount of the parent isotope from the measurements.

But Rb-Sr dating is  still susceptible to the amount of Rb and various Sr be changed by low grade metamorphosis after rock body formation.     You still need corroboration from other dating methods, or have ways for controlling for impact of low grade metamorphosis, such as good agreements in Rb-Sr dates across stratigraphically correlated rock formations over an area large enough to be implausible for the different rock samples to have undergone the same metamorphic history.