Analysis Of Inversion Method For Apatite Fission Track Annealing Process

The apatite fission track is a low-temperature thermochronology technique based on the radiation damage effect of U~23838 fission.The high-energy particle fragments generated by U~23838 fission can generate a radiation damage area in the apatite crystal.This trace of damage is the track.For different apatite crystals,the length of the track produced by fission is different.As the new track continues to be generated,the previously generated track will be shortened by the annealing effect,and two important factors affecting the annealing efficiency are temperature and time.The track distribution in apatite records the thermal evolution history of the apatite.By counting the number of tracks in the apatite particles,combined with the decay law of radioactive elements,the track age can be calculated,combined with the track length and the apatite fission track annealing kinetic model,the thermal history experienced of apatite can be reversed.Due to the low sealing temperature of apatite,it can effectively reflect the process of shallow crust change and the influence of surface conditions,so that the apatite fission track technology can quantitatively analyze the thermal events and the characteristics of velocity and spatial distribution in erosion events occurring in the shallow crust.Therefore,the apatite fission track can be widely used in recent years in the study of orogenic belt exhumation,topography and sedimentary basin evolution.At present,the main method for inversion of apatite fission track data is Monte Carlo method.In the process of inversion,different sampling algorithms have a great influence on the accuracy and efficiency of Monte Carlo method.At the same time,the inversion effect is explained by three parameters:GOF value(simulated track age compliance),K-S test value(track length distribution conformity)and time.The comparison shows that the GOF and KS values of the traditional Monte Carlo method are 0.89 and 0.86,and the Markov chain Monte Carlo(MCMC)algorithm are 0.96 and0.92.In the meantime,the former takes 15min,and the latter takes 21min,which indicates that the Markov chain Monte Carlo method is superior to the traditional Monte Carlo method in accuracy,and is slightly inferior in efficiency.