The increasing frequency and complexity of public health emergencies underscore the urgency of building an efficient emergency management system. This study focuses on Guangdong Province’s public health emergency management policies, employing Rothwell and Zegveld’s policy tool classification framework (demand-type, environmental-type, supply-type) combined with Guth’s time-risk management model (pre-event, during-event, post-event). Using text mining and quantitative content analysis methods, it conducts a two-dimensional cross-examination of policy documents from the Guangdong Provincial Health Commission (318 texts, screened to 178, yielding 843 effective tool units) and the Guangdong Provincial Center for Disease Control and Prevention (2 texts, 223 tool units). Results show that environmental-type tools dominate the policy mix (76.04%), with assessment and evaluation (33.57%) and legal norms (21.71%) as core elements, reflecting a regulatory and institutional orientation; supply-type tools account for 15.41%, while demand-type tools are minimal at 8.52%, indicating weak social participation mechanisms. In the temporal dimension, pre-event stage tools are most prevalent (43.40%), followed by during-event (35.85%), and post-event the least (20.72%), revealing a preventive bias but insufficient recovery support. Institutional comparisons reveal that Health Commission policies are comprehensive, covering multiple tools at all stages with an emphasis on top-level governance; CDC policies are specialized, focusing on drills and training (approximately 143 mentions) and prioritizing technical preparation. The study recommends strengthening post-event recovery policies and demand-side tool applications to enhance system resilience. This analysis uncovers the hierarchical logic of Guangdong’s public health emergency governance, providing an evidence base for regional governance optimization.