Accurate representation of nitrous acid (HONO) chemistry is critical for modeling atmospheric oxidation capacity and secondary pollutants like ozone (O₃) in China, yet current air quality models systematically underestimate HONO. To address this deficiency in the Community Multiscale Air Quality (CMAQ) model, we integrated five additional heterogeneous reactions: heterogeneous NO₂ reactions on ground/aerosol surfaces, particulate nitrate photolysis, NO_x oxidation, and acid displacement reactions. These updates were evaluated against wintertime HONO observations from the Lin'an Regional Atmospheric Background Station (LABS) in the Yangtze River Delta (YRD). The revised model reduced HONO underestimation dramatically, improving normalized mean bias from −90.0 % to −34.4 %. The simulation results demonstrated that ground-surface heterogeneous reactions dominated overall HONO production (45.4 %), peaking at night (65.3 %), while daytime formation was primarily driven by acid displacement (53.3 %). The enhanced HONO simulation amplified atmospheric oxidation capacity, increasing hydroxyl (OH) and hydroperoxyl (HO₂) radical concentrations by 87.6 % and 172.6 %, respectively. Consequently, O₃ peak simulations improved by 6.0–17.0 %, significantly reducing model bias (NMB: −8.9 % to −2.0 %) and better capturing pollution episodes. The model's enhanced representation of HONO formation significantly reduced the discrepancy between simulated and observed data, underscoring the critical importance of comprehending and accurately modeling HONO in the study of secondary pollution.