Continuous Diurnal–Nocturnal Observation of Atmospheric Aerosols: Mechanisms, Challenges, and the Prospects of AI Empowerment

There exist significant diurnal asymmetries in the physicochemical properties of atmospheric aerosols and their environmental and climatic effects. Such asymmetries stem from the diurnal variations of emission sources, the fundamental transformation of the thermal structure of the planetary boundary layer, and the alternating dominance of photochemical and heterogeneous chemical processes. However, traditional observational paradigms face enormous technical bottlenecks at night, resulting in a “half-day blind spot” in understanding the aerosol life cycle. This review systematically summarizes the capabilities and limitations of ground-based in-situ observations, active remote sensing, satellite measurements, and numerical models in revealing the mechanisms of diurnal aerosol variations. We point out that the current technical system suffers from a “diurnal–nocturnal separation” issue: ground-based point observations lack vertical dimensionality, lidar signal-to-noise ratio degrades during daytime, and passive satellite remote sensing fails at night due to the absence of light sources. This separation limits our ability to accurately assess the net radiative forcing of aerosols and their feedback effects on the evolution of the planetary boundary layer. Finally, we prospectively demonstrate the revolutionary potential of artificial intelligence (AI) technologies, particularly physics-informed machine learning and deep learning, in fusing multi-source heterogeneous data, reconstructing all-weather three-dimensional aerosol fields, and identifying key processes. By constructing an “AI-defined” intelligent observational network, we expect to achieve a cognitive leap from “phenomenon” to “mechanism” in aerosol research, providing unprecedented scientific support for climate prediction and pollution control.