Influence of Karst Landscape on Planetary Boundary Layer Atmosphere: A Weather Research and Forecasting (WRF) Model–Based Investigation

Abstract

Abstract Karst hydrology provides a unique set of surface and subsurface hydrological components that affect soil moisture variability. Over karst topography, surface moisture moves rapidly below ground via sink holes, vertical shafts, and sinking streams, reducing surface runoff and moisture infiltration into the soil. In addition, subsurface cave blockage or rapid snowmelt over karst can lead to surface flooding. Moreover, regions dominated by karst may exhibit either drier or wetter soils when compared to nonkarst landscape. However, because of the lack of both observational soil moisture datasets to initialize simulations and regional land surface models (LSMs) that include explicit karst hydrological processes, the impact of karst on atmospheric processes is not fully understood. Therefore, the purpose of this study was to investigate the importance of karst hydrology on planetary boundary layer (PBL) atmosphere using the Weather Research and Forecasting Model (WRF). This research is a first attempt to identify the impacts of karst on PBL. To model the influence of karst hydrology on atmospheric processes, soil moisture was modified systematically over the Western Kentucky Pennyroyal Karst (WKYPK) region to produce an ensemble of dry and wet anomaly experiments. Simulations were conducted for both frontal- and nonfrontal-based convection. For the dry ensemble, cloud cover was both diminished downwind of karst because of reduced atmospheric moisture and enhanced slightly upwind as moist air moved into a region of increased convection compared to control simulations (CTRL). Moreover, sensible (latent) heat flux and PBL heights were increased (decreased) compared to CTRL. In addition, the wet ensemble experiments reduced PBL heights and sensible heat flux and increased cloud cover over karst compared to CTRL. Other changes were noted in equivalent potential temperature (θe) and vertical motions and development of new mesoscale circulation cells with alterations in soil moisture over WKYPK. Finally, the location of simulated rainfall patterns were altered by both dry and wet ensembles with the greatest sensitivity to simulated rainfall occurring during weakly forced or nonfrontal cases. Simulated rainfall for the dry ensemble was more similar to the North American Regional Reanalysis (NARR) than CTRL for the nonfrontal case. Furthermore, the initial state of the atmosphere and convective triggers were found to either enhance or diminish simulated atmospheric responses.