The Data Assimilation Research Testbed: New algorithms and applications

NCAR's Data Assimilation Research Testbed (DART) provides a cutting-edge facility to perform ensemble data assimilation with support for a wide range of models, conventional and novel observations, algorithms, and diagnostic tools. Several research projects are presented that demonstrate the effectiveness and flexibility of the DART system to produce high-quality products. First, the Weather Research and Forecasting (WRF) model is used with DART to perform a CONUS domain mesoscale ensemble analysis with continuous cycling, providing initial conditions for regional convection-resolving ensemble forecasts. Second, a multi-year, loosely coupled DART atmosphere/ocean assimilation with the Community Atmosphere Model (CAM) and the Parallel Ocean Program (POP) has been completed to provide initial conditions for ensemble decadal ocean predictions. Some challenges of ocean data assimilation are presented, including the advantages of coupled ocean-atmosphere assimilations. Third, the impact of GPS radio occultation, and environmental satellite microwave and IR soundings of water vapor and temperature on analyses and forecasts of tropical storm tracks and intensities will be shown.

A new rank histogram algorithm for computing ensemble increments is introduced, expanding the number of filtering algorithms for any DART-compliant model to six. New tools are presented to explore and compare the performance of the assimilation configurations and observations.