Monitoring Fog Evolution of Air Quality in Central Taiwan Mountain Area Using Air Quality Boxes

Increased industrialization has led to heightened exposure to poor air quality, raising the risk of cardiovascular diseases. Forest environments are increasingly valued for their ability to improve human mental and physical health by releasing phytoncide and reducing air pollutants. Xitou Experimental Forest of National Taiwan University (23.40°N, 120.47°E, 1178 m a.s.l.), a cloud forest in central Taiwan, located in a valley linked to industrial and metropolitan areas to the northwest, experiences air pollutant transport influenced by land-sea and mountain-valley breezes. These local circulation patterns bring urban air pollutants inland during the day, causing higher daytime concentrations compared to nighttime levels. To evaluate the contributions of physical transport and chemical reactions across different seasons, we applied five home-built Air Quality Box (AQB) systems along the valley. Each AQB integrates low-cost sensors to monitor ambient gaseous pollutants (CO, NO, NO₂, O₃, SO₂, CO₂, and non-methane hydrocarbon), ambient particle number size distribution (0.38-17 μm diameter), and meteorological parameters (temperature, relative humidity, and pressure). The particle size distribution shifts toward larger sizes with elevation, driven by hygroscopic growth as relative humidity increases during parcel ascent. Aerosols might act as cloud condensation nuclei, forming fog droplets predominantly around 5 μm in diameter in the Xitou area. With Mie scattering calculations, the extinction effect of aerosols and visibility in the study area can be estimated. CO concentrations, a marker of local pollutant transport, increase with the development of sea breeze and valley wind but decrease as the cleaner mountain wind prevails. Seasonal variations show that the mountain wind develops earlier in winter than in summer, leading to earlier pollutant reductions. Furthermore, agglomerative hierarchical clustering of diurnal CO patterns shows how pollutant concentrations rise with the development of valley winds and decrease with mountain wind onset around 17:00. These results demonstrate the utility of AQBs in providing high temporal-spatial resolution data to analyze complex transport dynamics and fog formation processes in mountain environments.