Abstract: In order to help understand land surface energy balance processes and improve the inversion accuracy of canopy component temperature, an extended CUPID model was proposed to simulate detailed canopy temperature distribution. The extended model can simultaneously simulate the shaded and sunlit soil temperature based on the difference of net radiation and evaporation rate between the shade and sunlit soil. The field-measured data on winter wheat and summer corn were used to validate the model. In winter wheat field, the mean absolute errors were a little higher(2.8 K and 2.4 K) than the mean errors(-1.5 K and -0.7 K) between simulated and measured soil temperatures for sunlit part and shaded part of soil respectively. In the corn field, the mean absolute errors were much higher(3.8 K) than the mean errors(-0.5 K) between simulated and measured soil temperatures. As a whole, these results showed that the extended CUPID model could correctly simulate the spatial distribution and diurnal variation of the soil component temperature. The model can be applied to component temperature inversion and agricultural drought monitoring.