Continuous mobile sensing applications (CMSAs) run 24 hours in the background, consuming a significant amount of energy. Due to CMSA's dynamic behavior according to the users' context, it is very difficult to predict the energy consumption of CMSA. User trace-based solutions have been proposed to effectively estimate the energy consumption of CMSA, but they suffer burdensome user trace collection. We propose Scenethesizer, a scenario-based energy estimation system, which generates and augments the user traces for unseen application usage scenarios and estimate the energy consumption of CMSA based on the generated user traces.