<p>Addressing climate change is a major challenge facing human society in this century. To reduce the frequency of extreme weather events, governments around the world are actively implementing emission reduction policies, and public attention to carbon reduction is also increasing. Against this backdrop, this paper divides industries related to low-carbon transition into high-carbon and low-carbon sectors. By comprehensively applying methods such as TVP-VAR-SV, TVP-VAR-DY, wavelet coherence analysis, and partial wavelet coherence analysis, it explores the dynamic evolution of the relationship between climate policy uncertainty (CPU), public attention to carbon reduction (CRE), and extreme risk resonance (ERR) in low-carbon transition-related industries. Furthermore, the study employs the DCC-GARCH model to more accurately assess the dynamic correlation between high-carbon and low-carbon industries. The study mainly reaches the following conclusions: First, there are significant differences in the time-varying relationship among CPU, public attention to carbon reduction, and extreme risk resonance in low-carbon transition-related industries, with the response results of extreme risk resonance in high-carbon and low-carbon industries being noticeably different. Second, within the entire system, CPU mainly acts as a risk receiver, while public attention to carbon reduction functions as a risk spillover source, and the degree of spillover between extreme risk resonance in high-carbon and low-carbon industries also differs. Third, under shocks at different policy time points, when extreme risk resonance in high-carbon and low-carbon industries is affected by CPU and public attention to carbon reduction, the resulting impacts are significantly heterogeneous. In the long term, however, the synergy between CPU and public attention to carbon reduction demonstrates considerable independence.</p>