Sensitivity experiments with a high resolution data assimilation scheme

The Oklahoma Advanced Regional Prediction System (ARPS) Data Assimilation System (ADAS) has been implemented over the complex topography of northwest Utah and the Great Basin at near-real time, and at very high resolution. ADAS is a fully 3-dimensional analysis system which utilizes the Bratseth Method of successive corrections as the objective analaysis technique. Hourly analyses are run using the Rapid Update Cycle as the first guess field, modified by local surface observations, upper air soundings, radar data in NIDS format, wind profiler data and acars observations. The Intermountain Precipitation Experiment (IPEX) conducted during February 2000 provides a large, high resolution data set for a variety of winter storm events over northwest Utah. The data will be analyzed using ADAS in post-event research mode in order to improve the analysis of mesoscale storm structure and evolution, and to assess the impact of additional local data on the ADAS analysis quality. Analyses run in near-real time will be compared to analyses completed using the standard local data in addition to the full IPEX data set for a selection of storm events. Analysis error will also be assessed by computing the RMS error between the operational analysis values and the IPEX data values, at the observation locations.