Data Assimilation Research

Vertical distribution of time mean error (bias) and RMS error for analyses of temperature (left) and moisture (right) produced using the Data Assimilation Research Testbed and the WRF regional model with a North American domain. Blue dashed curves show errors compared to withheld radiosonde observations when assimilating only the remaining radiosonde observations while the generally lower-error red curves result when GPS radio occultation observations are also assimilated. These results demonstrate the positive impacts of using GPS observations, and they prepared the way for ongoing assimilation of GPS observations from the newly launched COSMIC satellites.

Data assimilation is the process of merging data from observations with computer models. It can transform diverse and incomplete observations to gridded estimates that can be easily used and interpreted. The assimilation process also produces quantitative information on model error, forecast skill, and observational errors, all of which allows us to improve the models. Data assimilation is providing rapid advances in geophysical studies. The Data Assimilation Research Section (DAReS) of IMAGe performs fundamental research on ensemble data assimilation methodologies for application across a wide range of geophysical problems. DAReS develops and maintains the Data Assimilation Research Testbed (DART), a software facility for doing ensemble data assimilation. DAReS also provides support to a growing community of NCAR, university, and government laboratory collaborators who are interested in applying ensemble data assimilation methods.

DAReS supports three of NCAR's strategic priorities: "Developing community models," "Developing and providing advanced services and tools," and "Enhancing science education." The DART user community includes members from many NCAR divisions, more than a dozen universities, and several government labs. Within NCAR,

• Researchers in CGD are using DART/CAM (Community Atmosphere Model) to validate and improve climate models,
• MMM is using DART/WRF (Weather Research and Forecasting model) to assimilate radar observations for convective-scale and hurricane prediction research,
• ACD is using DART/CAM to assimilate observations of CO,
• COSMIC is using DART/WRF to assimilate GPS radio occultation observations,
• RAL is using DART/WRF to study boundary layer assimilation and modeling, and
• HAO is exploring the feasibility of space weather applications.

University groups are using both DART/WRF and DART/CAM, and several researchers have incorporated their own models including hydrological models, small-scale tracer transport models, and ocean/atmosphere GCMs. Researchers at DOE/LLNL and NOAA/NSSL are also using DART products or software in their research. DAReS provides support for all these activities and uses feedback from users to develop more powerful and generic assimilation tools. DART has also been used to support graduate data assimilation classes at several universities. In FY 2007, one key focus will be collaborations with internal and external research groups studying carbon in the coupled climate system.

Fundamental data assimilation research focuses on advancing ensemble methods to make them more powerful and generic, capable of being effectively applied to many problems as nearly 'black-box' algorithms. Examples of recent advances that are now available in the DART framework are: hierarchical Bayesian filters that automatically and dynamically correct for ensemble sampling error; hierarchical Bayesian adaptive error correction methods that automatically detect and ameliorate the effects of model error; ensemble smoothers that use data from the past and the future to produce high-quality "reanalyses." FY 2006 has also seen development and deployment of a new scalable version of DART that runs efficiently on a variety of parallel computing platforms. In FY 2007, assimilation research will focus on improving methods for dealing with sampling error (a major concern for generic filtering algorithms) and using assimilation to determine the concentration, sources and sinks of atmospheric trace constituents.

This web page describes DAReS and the DART facility.

This project is made possible through NSF Core funding.