This paper employs a GARCH-MIDAS framework integrated with machine learning to investigate the impact of climate-related uncertainties on the volatility of the China's green bond market (GBM). By combining high-frequency financial data with multi-source, low-frequency climate uncertainty indicators, we examined how macro-financial conditions and climate risks jointly affect the dynamic changes of the GBM in China. Machine learning methods were used to identify and rank the key drivers of volatility. The research results indicate that traditional macro-financial variables remain the main determinants of the volatility in the green bond market, among which the impact of government bond yields is the most significant. Climate uncertainty information also has a significant impact on the volatility of green bonds. Moreover, incorporating climate uncertainty into the GARCH-MIDAS model significantly enhances its explanatory power, highlighting the importance of considering mixed-frequency risk factors in understanding China's green bond market dynamics. These findings underscore the crucial role of climate uncertainty in green bond pricing and indicate that combining machine learning with mixed-frequency volatility modeling can provide a more comprehensive framework for understanding the dynamics of the green bond market.