Needless to say, radiometric dating is not easy. It requires very careful measurement and analysis technique, as well as careful analysis of the assumptions, about such things as the purity and composition of the samples, to do this effectively. There are also assumptions about the accuracy of the equipment involved, and, of course, assumptions about the fundamental physics involved. The findings of the scientific community, on such things as the age of the Earth and the dates of various epochs (e.g. Permian, Triassic, Cretaceous, Paleogene) are the result of decades of measurement and analysis by hundreds of people.
== Key implausible assumptions ==All scientific investigations involve three general types of asumptions:#The scientists doing the investigation know how to perform the investigation—in short, they know what they are doing. If one is investigating temperature patterns to draw conclusions and predictions about weather, this means that the investigators understand the subject matter, and understand the connection between temperature and weather patterns.#The equipment works, is properly maintained, and is properly calibrated. That is, the thermometers work properly.#The fundamental scientific principles are understood. This means that it is known how the coefficient of thermal expansion works.
There are a number In the case of implausible assumptions involved in radiometric dating with respect , the first problem is by far the hardest. It takes years of careful observation and reasoning, by many scientists, to long time periodsreach a consensus about what actually happened and when.
=== Initial First assumption: initial quantities === One key assumption is that the initial quantity of the parent element can be determined. With uranium-lead dating, for example, the process assumes the original proportion of uranium in the sample. One assumption that can be made is that all the lead in the sample was once uranium, but if there was lead there to start with, this assumption is not valid, and any date based on that assumption will be incorrect (too old). Fortunately, isotopic analysis of the lead can mitigate the uncertainty, as can measurements of many samples in many contexts.
In the case of carbon dating, it is not the initial quantity that is important, but the initial ratio of C<sup>14</sup> to C<sup>12</sup>, but the same principle otherwise applies.
Recognizing this problem, scientists try to focus on rocks that do not contain the decay product originally. For example, in uranium-lead dating, they use rocks containing [[zircon]] (ZrSiO<sub>4</sub>), though it can be used on other materials, such as [[baddeleyite]].<ref>{{Cite journal|doi=10.2113/104.1.13 |title=SHRIMP baddeleyite and zircon ages for an Umkondo dolerite sill, Nyanga Mountains, Eastern Zimbabwe |year=2001 |first=M.T.D. |last=Wingate |journal=South African Journal of Geology |volume=104|issue=1 |pages=13–22}}</ref> Zircon and baddeleyite incorporate uranium atoms into their crystalline structure as substitutes for [[zirconium]], but strongly reject lead. Zircon has a very high closure temperature, is very chemically inert, and is resistant to mechanical weathering. For these reasons, if a rock strata contains zircon, running a uranium-lead test on a zircon sample will produce a radiometric dating result that is less dependent on the initial quantity problem.
=== Rate Second assumption: proper calibration ===This is fairly easy. How to calibrate radiation detectors is a problem that has been worked on since the days of Pierre and Marie Curie. === Third assumption: understanding of the rate of decay ===This one is also easy.
Another assumption is that the rate of decay is constant over long periods of time. Radiometric dating requires that the decay rates of the isotopes involved be accurately known, and that there is confidence that these decay rates are constant. Fortunately, this is the case. The physical constants (nucleon masses, fine structure constant) involved in radioactive decay are well characterized, and the processes are well understood. Careful astronomical observations show that the constants have not changed significantly in billions of years—spectral lines from distant galaxies would have shifted perceptibly if these constants had changed. In some cases radioactive decay itself can be observed and measured in distant galaxies when a supernova explodes and ejects unstable nuclei.