The Impact of Unmanned Aircraft System Observations on Convective Initiation Along a Boundary in Numerical Weather Prediction

Due to their ability to provide meteorological observations in challenging and undersampled environments, Unmanned Aircraft Systems (UASs) have the potential to enhance Numerical Weather Prediction (NWP). An Observing System Simulation Experiment (OSSE) is used herein to evaluate the potential impact of observations from this new platform on NWP output. An OSSE is used to simulate observations collected using observing platforms--such as UASs.

The severe weather outbreak that occurred in south-central Oklahoma on 9 May 2016 is being used for this study. Severe thunderstorms developed along and ahead of a dryline during the afternoon and produced 12 tornadoes in Oklahoma including an EF-4 tornado that caused one million dollars in damage and one death. The effectiveness of use of UAS observations to simulate convective initiation along this dryline is considered herein.

North American Mesoscale Forecast System Analyses (NAM-ANL) data were obtained from the National Centers for Environmental Information (NCEI). These data were ingested into the Weather Research and Forecasting (WRF) Model to create a base state atmosphere. The simulations utilized an outer grid spacing of 6 km and an inner grid spacing of 2 km. In order to create synthetic UAS observations, a second model run was conducted at a higher resolution. Simulated UAS flights were used to collect state atmospheric variables including temperature, pressure, dew point, and wind speed and direction as the UAS traversed the boundary. The types of flights included a single UAS mission and a multiple UAS mission. Alternate flight path configurations and altitudes were evaluated to determine the most effective UAS sampling method. After collecting synthetic UAS observations, the WRF Data Assimilation System (DAS) was utilized to assimilate these data into the lower resolution model run to examine the impact of synthesized UAS observations on the NWP model forecast. Lower resolution output with and without simulated UAS observations are compared.