Climate change is a long-term, global, environmental damaging, technology-dependent and uncertain issue. Despite substantial work on both model approaches and applications, there are several limitations in current researches: first, technological maturity, crowd-out effect and spill-over effect of R&D are important aspects of endogenous technological advances, but they were considered in current studies. Second, non-market damage is still a tough nut to crack for climate damage assessment, but not enough attention is paid to the adaption which proves to be equally important for mitigating the climate change. Finally, little efforts were taken to the pure time preference rates and learning rates, whose uncertainties will to a large extent affect the allocation of utility among different generations and the effectiveness of technical learning(including both learning-by-doing and learning-by-searching effect), respectively.