Uncertainty Of Model Parameter In WS* Nuclear Mass Model

The experimental measurement and theoretical calculation of nuclear mass are always the hot topics in nuclear physics.Up to now,the masses of measured nucle-i in the experiment is about 2400,and there are about 5000 nuclear masses waiting for experimental measurements according to model predicts.There are many kinds of models to calculate the nuclear mass,such as the microscopic model,the macroscopic-microscopic model,the phenomenological model,the shell model,the nucleon inter-action model and so on,which one of the most representative is the macroscopic-microscopic model(such as Weizacker-Skyrme nuclear mass model).The 2149 known nuclear mass of the rms deviation calculated by the WS*model is 0.441MeV,at the same time,it has only 13 parameters.Considering the high accuracy and less com-putation time of the WS*model,it is meaningful to study the uncertainty of parameters based on the WS*model.The investigation of the influence of parameters on the mass predictions and the correlation among the parameters is of great academic significance to further improve the nuclear mass formula.In this paper,we first test the predictive power of the nuclear mass formulas.We have calculated the masses of the 270 new nucleus in AME2016 which is the unknown in AME2003 by comparing the calculated results of the FRDM model,the WS model and the WS*model.We found the rms deviation of the WS*model is the smallest one with a valle of which indicates that the predictive power of the WS*model is strong.On the other hand,we study the uncertainty of the WS*model parameters and the correlation between the model parameters.Through studying the thirteen parameters of the model one by one,we found that the uncertainty of pairing correction coefficien-t apair,depth of single particle potential V0 and spin orbit potential coefficient V0 is relatively large.Through studying the variation of the rms deviation of 2149 known nuclear masses with any two parameters,we calculated the pearson correlation coef-ficient between any two model parameters and we found that the correlation between macroscopic parameters is relatively more strong and the correlation between macro and micro parts is relatively weak and the correlation between microcosmic parameters is relatively weak.At the same time,we study the statistical error of nuclear mass cal-culation.We found the deviations between experimental measurements of nuclear mass in AME2003 and in AME2016 and the theoretical predictive value of the WS*model all fall within the range of error,which indicates that the standard deviation of the 13 parameters is reliable.For neutron-rich and superheavy regions and the magic nucle-us,their statistical error is relatively large.For the extremes neutron-rich nucleus,there are two parameters related to symmetry energy(csym and ?)cause great quality uncer-tainty.The influence of r0 on light nucleus and magic nucleus is relatively large.More-over,for relatively heavy deformed nuclei,the influence of g1 and g2 is strong.Which indicates that the most sensitive parameters are different for different nuclei.These results are helpful for further study of the statistical uncertainty of WS*nuclear mass formula.