Building State-of-the-Art Forecast Systems with the Ensemble Kalman Filter
Jeff Anderson
University Corporation for Atmospheric Research
ABSTRACT: The development of numerical weather prediction was one of the great scientific and computational achievements of the last century. The best present-day numerical weather prediction systems have evolved over decades and feature model-specific assimilation systems built with nearly a person century of effort. Anderson describes the development of a community software facility for ensemble Kalman filter data assimilation, the Data Assimilation Research Testbed (DART). DART can produce high-quality weather predictions but can also be used to build a comprehensive forecast system for any prediction model and observations. Anderson describes the basic ensemble Kalman filter and applies it to simple example problems. Anderson presents heuristic extensions to the basic algorithm that are essential for large applications in a historical context. An ensemble forecast system can estimate model parameters, guide general model improvement, evaluate the quality of existing observations, and inform the design of future observing systems. Examples of these capabilities are provided for a variety of geophysical applications.
