Abstract: As one of the most promising water-saving rice production technologies, the ground cover rice production system (GCRPS) has been found to save water application, increase soil temperature, and reduce nitrogen pollution and methane emission. However, the feasibility of CERES-Rice, a software package widely and successfully applied in the traditional paddy rice production system (TPRPS), for simulating the rice growth in the GCRPS still remains unknown and needs further research. Undoubtedly, it should be based on accurately quantifying the effect of soil temperature enhancement caused by the ground cover material (chosen as the plastic film in this study). Therefore, the objective of this study was to improve the two simulation models for both soil surface and subsurface temperatures in CERES-Rice through taking the effect of soil temperature enhancement by the film mulch into consideration. The simulation model of surface soil temperature (at the depth of 5 cm) was referred from other study for dry land crops, and the other one was from CERES-Rice for simulating the subsurface temperatures (at 10 and 20 cm, respectively) in the TPRPS. To justify and rectify the simulation models, we conducted a field experiment in Fangxian, Hubei, China (32°7′N, 110°42′E, altitude 450 m) from 2013 to 2014, covering two growth seasons of rice. Three treatments (named as TPRPS, GCRPSsat and GCRPS80%, respectively) were designed and replicated three times in 9 plots, each with an area of 9×10 m2. A seepage-proof material was laid around each plot to the depth of 80 cm to avoid lateral percolation between neighbor plots. Five soil beds (156 cm wide and 940 cm long) in each plot were built for planting rice, with the space of 26×18 cm2 and at a rate of two plants per hill. Small furrows (15 cm in width and depth) were dug around each soil bed. In the three replicated plots without plastic film for treatment TPRPS, a water layer of 2-5 cm in thickness was always maintained on the soil beds. In the three plots with plastic film for GCRPSsat, the root zone averaged soil water content was kept close to saturation by completely filling the furrows with water but without water layer on the soil beds. The remaining three plots with plastic film for GCRPS80% were managed as the same way as that for GCRPSsat before mid-tillering stage, and then transient irrigation was intermittently implemented through the furrows to keep the root zone averaged soil water content between 80% and 100% field water capacity. Among the two growth seasons, the experimental data obtained in 2013 and 2014 were used to rectify the simulation models and verify the rectified models, respectively. Based on the measured air temperatures, soil water contents, soil physical parameters and organic matter contents, and other related heat coefficients, the changing processes of soil temperature at the depths of 5, 10, and 20 cm in the two GCRPS treatments were simulated using the rectified models. The simulated and measured surface soil temperatures at 5 cm during both growth seasons were in good agreement, with the root mean squared error (RMSE) less than 1.8℃, normalized root mean squared error (NRMSE) less than 10%, and correlation coefficient (r) higher than 0.89 (P < 0.01). The simulated subsurface soil temperatures at 10 and 20 cm in 2013 or in 2014 were also within acceptable ranges, with RMSE < 3.2℃, NRMSE < 15%, and r > 0.65 (P < 0.01), respectively, between the measured and simulated values. The rectified models should be helpful to simulate the changing processes of soil temperature or soil heat transfer, and improve CERES-Rice for further evaluating rice growth in the GCRPS.