•We established single leaf area model of two broad-leaved species, Tilia amurensis and Betula platyphylla.•The general mode for Tilia amurensis, Betula platyphylla and Pinus koraiensis was to turn on the corresponding phenological events when the AAT reaches a certain threshold.•Compared with those with temperature as the independent variable, the phenology models with time as the independent variable had lower goodness of fit.•AAT-based models were used to establish a leaf area index prediction model that reflects the impact of abnormal climate and improves prediction accuracy and reliability. Global warming is causing substantially earlier spring leaf phenology in temperate‐zone trees. Understanding the drivers and mechanisms behind the observed plant phenological changes is important for predicting future phenological dynamics. Here, leaf phenological dynamic data (including number and area) observed in situ for three tree species (namely, Tilia amurensis, Pinus koraiensis and Betula platyphylla) on Changbai Mountain from 2017 to 2020 were analyzed. We found that none of the tree species showed active accumulated temperature (AAT) differences in phenological events and phase; however, differences in time (i.e., photoperiod) were observed and could be explained by AAT being the primary controller of plant phenology. Moreover, we developed process-based spring leaf phenology models that were fitted and validated using the leaf development process and time series. Compared with the time-based models, the phenology model with AAT as the independent variable is more robust in fitting and predicting leaf area and leaf number. Our novel findings provide evidence of AAT effects on leaf unfolding, whereby when the AAT reaches a certain threshold, the corresponding phenology will be triggered. Therefore, we used AAT-based models and plot survey data to simulate the leaf area index (LAI) dynamics of the tree species studied, which provides a feasible method to understand the complex processes scaling up from the plant and forest‐levels. Our study has provided a new answer to how temperature affects spring leaf phenology in temperate forests, and significantly improved the predictability for leaf development.