A physically based empirical model was established for the beginning and ending dates (BED) of limited temperatures.Data used were daily mean temperatures from 34 conventional stations in the Chongqing region,measured from 1961 to 2000.Using this model in combination with the distributed modeling results of monthly mean temperatures,we developed a method for creating a distributed simulation of the BED and durations of limited temperatures in the rugged terrain of Chongqing.Spatial distribution maps were generated at a resolution of 100×100 m for the BED and durations of the limited temperatures of 0,10,15 and 20.Simulations suggest that the distributed model is able to successfully reveal the macroscopic trends and local distributions.For the modeled BED,the mean absolute bias error is 0.6-2.4 d.Cross-validation and case-year tests show mean absolute bias errors of 1.3-2.8 d and 0.7-2.5 d,respectively.Comparison shows that the simulation results are consistent with the field investigations.This implies that the simulations are as good as observations,thus providing a viable technique for simulating mountainous heat resources using conventional meteorological data.