Methane hydrates (MHs) are deemed to be a potential energy source to meet the overwhelmingly growing demand for natural gas. The production of MHs involves many controlling processes such as heat transfer, fluid dynamics and hydrate reaction kinetics. Local thermal equilibrium (LTE) between solid and fluid phases was typically assumed when modeling the hydrate production, however, in many cases the local thermal non-equilibrium (LTNE) caused by the complex heat transfer processes at the pore scale cannot be neglected. In this work, a LTNE model was adopted to reflect the multiple heat transfer forms associated with hydrate dissociation and gas-liquid flow, particularly the solid-fluid interfacial convective heat transfer and the Joule-Thomson effect. The LTNE phenomenon and Joule-Thomson effect were hinted by a hydrate dissociation experiment carried out for model verification. The results show that the average estimation errors of temperature, cumulative gas and water production profiles between experiment and LTNE model were 7.70%, 6.63% and 4.02%, respectively, which are considerably better than those of the traditional LTE model (16.44%, 11.74% and 10.89%, respectively). A tenfold increase of interfacial convective heat transfer coefficient could reduce the local solid-fluid temperature difference by about 78.5%. A special focus was given to the Joule-Thomson effect which seemed to have less impact at the experimental condition in this work but become more significant when the gas saturation exceeds 40%. This work could add further insights into the thermal performance during hydrate dissociation in porous media, help to reveal the thermo-kinetic relationship at the pore scale and provide a more accurate estimation of the hydrate dissociation dynamics. It could be potentially employed to optimize the methane hydrate production strategies, minimize the risk of hydrate re-formation and maximize the overall energy recovery efficiency.