12B.3 Atmospheric Analysis Uncertainties over the Antarctic Region using an EnKF Method with the AMPS Model

Thursday, 2 July 2015: 8:30 AM
Salon A-5 (Hilton Chicago)
Christopher P. Riedel, University of Oklahoma, Norman, OK; and S. Cavallo

Atmospheric Analysis Uncertainties over the Antarctic Region using an EnKF Method with the AMPS Model

Christopher Riedel

The Antarctic and Southern Ocean are extremely difficult regions for numerical weather prediction (NWP) models in part due to the limited quantity and quality of observations for this remote region. From this a greater degree of uncertainty exits in accuracy of the analyses utilized for the model initialization, potentially impacting forecast skill. The Antarctic Mesoscale Prediction System (AMPS), which is a modified version of the Advanced Research Weather and Research Forecasting (ARW-WRF), is the only operational mesoscale NWP model used for the Antarctic region. We have developed an ensemble-based data assimilation in the AMPS model by utilizing an Ensemble Adjustment Kalman Filter (EAKF) technique within the framework of the Data Assimilation and Research Testbed (DART). We refer to this new numerical modeling system as Antarctic-DART. In this study, we evaluate experiments designed to isolate the extended range predictive skill of Antarctic-DART through improvements in the (1) assimilation of polar orbiting observations and (2) model physical parameterizations. We perform these experiments during the Concordiasi intensive observation period in September – November 2010, where special dropsonde observations provide an independent validation of Antarctic-DART performance throughout the depth of atmospheric columns.

We will begin our discussion by evaluating results from a control experiment where only conventional observations are assimilated, which are surface (METAR, AWS, ship), rawinsondes (wind, temperature and relative humidity), aircraft (ACARS), cloud satellite wind from geostationary satellites, and GPS (COSMIC) observations. The control experiment immediately revealed a large-scale upper-level circulation bias, namely that circulation around the Antarctic polar vortex is too weak in Antarctic-DART. In our first experiment, we additionally assimilated cloud satellite winds from polar orbiting satellites. However, this data was primarily located in inside the vortex, and hence did not have a large impact on the equator-to-pole temperature gradient controlling the strength of the larger-scale vortex. In the second experiment, we implemented a time-varying ozone distribution to more accurately reproduce the shortwave heating rates in and around the Antarctic ozone hole. This substantially reduced model bias at all vertical levels from the top of the model to as low as 700 hPa, however this did not improve model biases near the surface. Finally, we will examine the contributions of the geostationary satellite winds to the upper-level circulation bias.

- Indicates paper has been withdrawn from meeting
- Indicates an Award Winner