Abstract
Abstract. Recent years have seen the emergence of local air pollutant monitoring networks that feature close proximity to urban activities, higher requirement for temporal granularity, and improvisions in equipment hardware and installation conditions. These networks are intended for the pertinent monitoring and improvement of urban air quality, but potential technical issues may undermine their ability to serve such purposes. This study utilizes a minute-granularity network in a university campus, and designs and conducts a series of experiments on how it performs under practical scenarios, including response to sudden environment change, reflection of multi-scale influencing factors, usability of different baseline stations, and ability to detect local emission events. Statistical and signal-processing technics are applied for understanding these experiments. The results indicate the source of complexity in such networks, the preferred temporal granularity for capturing different temporal patterns, the necessary reserved time for mobile stations, and the sensor location requirement for monitoring local emission events etc. In practical terms, these results provide a large amount of information on the specific capabilities and limitations of a near-surface, high-granularity monitoring network in the urban environment, and what to consider and to expect as a designer or user of such a network. In scientific terms, it is a strong reminder of the significance of numerous uncertainty issues in similar empirical studies.