Abstract: To address the issue of inaccurate load extrapolation results caused by the subjective experience-dependent threshold selection using the graphical method in peak over threshold (POT) time-domain extrapolation, this study proposes a threshold optimization method based on the mean excess function-least squares method (MEF-LSM). Initially, the empirical mode decomposition (EMD) method is employed to decompose the non-stationary loads of the combine harvester during field operation into dominant loads and trend loads. Subsequently, the MEF method is employed to determine the candidate threshold interval for the dominant loads, and the relationship between the LSM characteristic parameters of different candidate thresholds and the extrapolation result validation indicators corresponding to different candidate thresholds is investigated, with the threshold corresponding to the minimum characteristic parameter selected as the optimal threshold. Then, extreme loads exceeding the threshold are extracted, and the dominant loads obtained from empirical mode decomposition are extrapolated. The distribution of extreme points is fitted using the generalized pareto distribution (GPD), and the extrapolated dominant loads and trend loads are linearly combined to obtain the time-domain extrapolated loads for the combine harvester. Ultimately, the time-domain extrapolation results obtained by the MEF-LSM are compared with those obtained by the mean excess function threshold selection method. The results indicate that the smaller the characteristic parameter, the higher the accuracy of the corresponding threshold extrapolation result. Compared with the threshold extrapolation result obtained by the mean excess function graphical method, the MEF-LSM method improves the validation metric R² by 1.23%, verifying the effectiveness of the MEF-LSM method. The research findings can serve as a reference for the compilation of load spectra for combine harvesters, providing a basis for accurate fatigue life prediction and reliability analysis of agricultural machinery equipment.