In the digital measurement of aircraft assembly, the accuracy of large-size measurement field construction is highly dependent on the stability of the reference position laid on the tooling. The position of the reference points of a large tooling is highly susceptible to thermal drift due to changes in ambient temperature, leading to a reduction in the accuracy or even failure of the measurement field. Therefore, this paper takes a combined large-scale tooling as an example, and constructs a numerical model for predicting the thermal drift of the reference points of a large-scale tooling under the non-uniform temperature field by collecting the measured temperature and coordinate data at the reference points of the tooling in the field; based on the large amount of thermal drift data obtained from the simulation of the aforementioned model, the surrogate model for the thermal drift of the tooling is constructed by using BP neural network; and finally, the measured and the surrogate model are compared and analyzed in terms of the drift data of the reference points temperature-coordinate. The results show that the average relative errors of the simulation results are all below 18%, and the average relative errors of the BP neural network results are all below 22%, realizing the effective prediction of the thermal drift of the reference points.