Driving factors of HONO and their synergy in Hohhot, a semi-arid city in northern China: Insights from interpretable machine learning approaches

Abstract

Nitrous acid (HONO) serves as the key source of hydroxyl radicals (OH) in the atmosphere, significantly impacting atmospheric photochemical reactions and air quality. However, the source and formation process of HONO are influenced by numerous factors, and their synergy leads to a limited understanding of its formation mechanisms. In this study, we utilized a random forest (RF) model to elucidate the nonlinear relationships between these driving factors and HONO concentrations in winter and summer. To gain further insights, we combined RF model with SHapley Additive exPlanations and Partial Dependence Plot methods to reveal the sources, driving factors, and their synergy. The findings suggest that HONO is primarily driven by both primary emissions and NO₂ transformation in winter, but by the heterogeneous transformation of NO₂ in summer. A potential connection exists between winter PM_2.5 pollution and HONO. By limiting primary emissions, HONO concentrations can be reduced, which in turn helps mitigate winter PM_2.5 pollution. To lower HONO levels in summer, a synergistic approach is needed to control both NO₂ and NH₃, particularly under high relative humidity. The findings presented herein enhance our understanding of the HONO formation mechanisms and provide a scientific basis for developing comprehensive air quality improvement strategies.