The Loess Plateau, a typical ecologically fragile region in China, is increasingly facing complex and intensified wildfire risks due to the combined impacts of climate change and human activities. Identifying the spatiotemporal patterns of wildfire hazard and assessing its potential ecological impacts are crucial for regional ecological security and fire prevention. Utilizing multi-period wildfire data from 2001 to 2024, this study systematically analyzed the temporal evolution of wildfire activity. A machine learning algorithm was employed to construct wildfire hazard models, which were then coupled with Ecosystem Service Value (ESV) to conduct an ecological risk assessment. The results revealed significant spatiotemporal heterogeneity in wildfire hazard, with notable increasing trends observed during summer, autumn, and the non-fire prevention season, highlighting a newly identified characteristic of escalating fire risk outside the traditional fire prevention period. High-hazard zones were more extensive and exhibited continuous belt-like distributions in spring, winter, and the fire prevention season. The driving mechanisms of hazard factors varied across periods, clarifying the seasonal shift in the contributions of meteorology, vegetation, and human activities. The spatial distribution of ESV exhibited a gradient pattern, with higher values in the southeast and lower values in the northwest. Spring, winter, and the fire prevention period were identified as critical phases for ecological risk management, characterized by highly continuous high-risk areas and significantly higher projected ESV losses compared to other periods. The findings provide scientific support for developing spatiotemporally differentiated fire prevention strategies and for maintaining ecological security in the Loess Plateau.