Mid-Atlantic nocturnal low-level jet characteristics: a machine learning analysis of radar wind profiles

This paper introduces a machine-learning-driven approach for automated nocturnal low-level jet (NLLJ) identification using observations of wind profiles from a radar wind profiler (RWP). The work discussed here is an effort to lay the groundwork for a systematic study of the mid-Atlantic NLLJ's formation mechanisms and their influence on nocturnal and diurnal air quality in major urban regions by establishing a general framework of NLLJ features and characteristics with an identification algorithm. Leveraging a comprehensive wind profile dataset maintained by the Maryland Department of the Environment's RWP network, our methodology employs supervised-machine-learning techniques to isolate the features of the southwesterly NLLJ because of its association with pollution transport in the mid-Atlantic states. This methodology was developed to illuminate spatiotemporal patterns and physical characteristics of NLLJ events to study their role in planetary boundary layer evolution and composition. This paper discusses the construction of this methodology, its performance against known NLLJs in the current literature, intended usage, and a preliminary statistical analysis. The results from this analysis have yielded a total of 90 southwesterly NLLJs from May-September of 2017-2021, as captured by the RWP stationed in Beltsville, MD (39.05 degrees N, 76.87 degrees W; 135 m a.s.l.). A composite analysis of 90 jets reveals that the mid-Atlantic NLLJ is characterized by a core wind speed exceeding 10 m s-1 at altitudes typically between 300-500 m above ground level, with maximum wind speeds occurring between 3-6 h after sunset. The jets show consistent wind direction from the southwest but transition from more southerly- to more westerly-dominated with increasing altitude and time after sunset. We hope our study equips researchers and policymakers with further means to monitor, predict, and address these nocturnal dynamics phenomena that frequently influence boundary layer composition and air quality in the US mid-Atlantic and northeastern regions.