Analysis of the AMPS-Polar WRF Boundary Layer at the Alexander Tall Tower! site on the Ross Ice Shelf

Flight operations on the Ross Ice Shelf rely on accurate forecast aided by the AMPS-Polar WRF. To improve the performance of the model planetary boundary layer (PBL), this study examines two datasets containing vertical meteorological data on the western Ross Ice Shelf and compares them against the AMPS-Polar WRF. The first dataset contains nearly two years of meteorological data provided by the Alexander Tall Tower! (ATT) automated weather station from March 2011 – July 2012. Maintained by a joint team at the Space Science and Engineering Center (SSEC) and the Cooperative Institute for Research in Environmental Sciences (CIRES), the ATT provides instantaneous observations at 10-minute intervals at various levels up to 30 m. The second dataset comes from the same researchers from CIRES who conducted a field campaign from January 13th–26th, 2014 using aerial Small Unmanned Meteorological Observer (SUMO) vehicles to measure PBL conditions near the ATT site. The ATT provides temperature, moisture, wind speed, wind direction, pressure, longwave radiation, and shortwave radiation measurements while the SUMO flight data contains observed temperature and relative humidity, and calculated wind speed and wind direction. For the analysis of the ATT data, the 5-km AMPS-Polar WRF data is run daily at 00 UTC and 12 UTC. Each model run is given a 12-hour spin-up and the subsequent 12-23 forecast hours are concatenated to create a continuous hourly forecast record. For the SUMO analysis, flight data and ATT data are compared against vertically and horizontally interpolated 3-km AMPS-Polar WRF data up to 800 m above the surface. On a synoptic scale, European Centre for Medium-Range Forecasting ERA-Interim and AMPS-Polar WRF are utilized to identify possible sources of error in the AMPS-Polar WRF PBL.From the combined analysis from the ATT – AMPS climatology and the SUMO case studies, the most common errors are: a systematic dry bias and high wind speed bias in AMPS-Polar WRF. The ATT climatology shows a 10% dry bias year round that is highest when wind speeds exceed 15 m s-1. Diurnally, the dry bias is most evident at night during the summer when AMPS-Polar WRF and observations diverge in relative humidity with observations showing an increase while AMPS-Polar WRF decreases. During the SUMO case studies, four of the six days exhibited dry biases. AMPS-Polar WRF favors more dry continental air masses than the ECMWF ERA-Interim. Even when AMPS-Polar WRF and ERA-Interim agree on the strength of the katabatic flow off the Transantarctic Mountains, AMPS-Polar WRF appears to be drawing drier air down from the Antarctic Plateau. The wind speed transition range for stability errors is 4-8 m s-1, where AMPS-Polar WRF goes from underestimating to overestimating inversion strength. AMPS-Polar WRF appears to overestimate the wind speed during periods stratifying conditions at the ATT by 1-2 m s-1. The ATT analysis shows that AMPS-Polar WRF has fundamental issues with handling stability as a function of wind speed. A high wind speed bias also appears four of the six days during the case studies and during winter 2011. The ERA-Interim synoptic charts during the case studies suggest the wind speed errors are likely related to katabatic flow placement errors, overestimating the strength of cyclones over the Ross Sea, and overestimating barrier wind jet strengths. The current goal is to modify inadequate model physics to improve the accuracy of the PBL