Spatiotemporal characteristics of deep convection initiation in the Central United States

Stephen A. Shield,Adam L. Houston
DOI: https://doi.org/10.1002/joc.8472
2024-05-14
International Journal of Climatology
Abstract:The spatiotemporal characteristics of thunderstorm initiation is important to better understand thunderstorm initiation and the mechanisms responsible for it. This study examines thunderstorm initiation over the Central United States from a climatological perspective using a large sample of thunderstorm initiation points and utilizes spatial statistical and machine learning analysis methods. Results show the spatial favourability and spatial variability in the diurnal cycle of thunderstorm initiation. This research examines the spatiotemporal characteristics of deep convection initiation (DCI) using a dataset of approximately 182,000 instances of DCI occurring over an 11‐year period in the Central United States. Spatial statistical analysis reveals differences in the frequency of DCI occurrence across the study area. Favoured areas of DCI are most common across areas of higher terrain and in proximity to the Gulf Coast, while DCI occurs less frequently over the northern Great Plains. Spatial differences in the degree of interannual variability of DCI are also noted, with higher (lower) interannual variability generally noted in areas of less (more) frequent DCI occurrence. Unsupervised machine learning is utilized to identify spatial variability in the diurnal cycle of DCI. The data and methods in this study are able to reliably identify more obvious differences in diurnal time series of DCI occurrence, while more subtle differences such as those resulting from localized or weaker mechanisms are identified less reliably. However, from these results it is obvious that spatial variability in the diurnal time series of DCI exists.
meteorology & atmospheric sciences
What problem does this paper attempt to address?