Directorate

• Contents
• EXECUTIVE SUMMARY
• 1. SCOPE OF THE PROGRAM
• 2. IMPROVING OUR UNDERSTANDING OF EARTH-SYSTEM PROCESSES
• 2.1 Scale Interactions Associated with Precipitating Convective Systems
• 2.2 Understanding the Dynamics and Predictability of Weather Systems on Time Scales of 0-48 h
• 2.3 Fundamental Research on Precipitation Processes, with an Emphasis on their Improved Representation in Numerical Models
• 2.4 Understanding the Dynamics, Physics and Chemistry of Multi-Scale Atmospheric Chemical Constituent and Particulate Transport, Dispersion and Transformations
• 2.5 Fundamental Research in Boundary Layers and their Interfacial Interactions
• 3. ADVANCING THE SCIENCE OF ATMOSPHERIC PREDICTION
• 3.1 Advancing Short-Range Weather Prediction
• 3.2Advancing Data Assimilation Research
• 3.3 Improving Forecasting of Pollutants and Hazardous Materials
• 4. ADVANCED COMMUNITY RESEARCH TOOLS
• 4.1 WRF Community Model
• 4.2 Community Data Assimilation Techniques
• 4.3 Nested Community Climate Model
• 5. SUMMARY
• Acronym List
• Print Version
• Download Printable PDF

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Mesoscale & Microscale Meteorology Division Science Plan:

3.2 Advancing Data Assimilation Research

Goal: To explore advanced assimilation techniques, including Four-Dimensional Variational (4D-Var) and Ensemble Kalman Filter (EnKF) approaches, aimed at developing a unified WRF data assimilation system and with a particular focus on assimilating remotely sensed radar data at convective/mesoscales.

Our work on advanced data assimilation has both research and applied aspects. At the current stage of the program, we have an active program of research into a number of aspects of mesoscale data assimilation, including 4D-Var and EnKF approaches. As this work progresses, we shall make the advances available to the university community and operational centers. Such advanced data assimilation systems will enable MMM to address a number of key research questions aimed at improving the prediction of high-impact weather events, such as severe convection, and hurricanes. Our plans include:

• Further development of advanced assimilation techniques applicable to the complex, nonlinear convective/meso-scale regime; and
• Application of these techniques in simulated and real-world experiments to assess and maximize the impact of both current and future observing networks.

Both plans are consistent with THORPEX goals for advancing data assimilation.

A major goal of our data assimilation efforts is to extend the current WRF 3D-Var capability to include the time dimension (4D-Var). This task will combine research studies of the role of flow-dependent, dynamically consistent forecast error covariances, together with the impact of 4D-Var's capability to analyze the time tendency of high-frequency observations, for example, radiometers, radar, and radiances.

The EnKF is a promising approach to developing a combined ensemble forecast and data-assimilation system. In essence, one begins with an ensemble of analyses and makes an ensemble of short-range forecasts (using the full nonlinear forecast model) to the time of the next available observations. This ensemble of forecasts is used to estimate the forecast covariances required to assimilate the new observations via the standard Gaussian formalism of the Kalman filter. Each ensemble member is then updated given the new observations by assimilating a set of perturbed observations (that is, the actual observations plus noise consistent with the observational uncertainty). This approach shares with 4D-Var the benefit of flow-dependent forecast covariances, but, unlike 4D-Var, it does not require the linearized or adjoint versions of the forecast model, nor does it require the off-line estimation of first guess forecast error covariances. The ensemble Kalman filter also has the attractive feature of providing a short-range ensemble forecast and of “initializing'' the ensemble members for longer-range ensemble forecasts. Our previous significant initial work in this area provides a good foundation for our planned further work.

The optimal use of existing observation networks is a key aspect of data assimilation research. With regard to the prediction of severe convection, Doppler radar data is a primary source of information. Efforts will focus on the optimal preprocessing, quality control, and thinning of radar radial velocity and reflectivity observations. This will be achieved via collaboration with radar data expertise in EOL, RAL and the wider research community, and will complement our more general research into assimilation techniques. Satellite observations, especially microwave and infrared radiances, also are an important source of information for NWP, especially at global and synoptic scales. We plan to enhance the existing capabilities of the WRF system for assimilating such observations. Because satellite radiance data assimilation techniques are already well established elsewhere, our limited expertise in this area will be expanded and supplemented via collaborations with community satellite data expertise (e.g., CSU/CIRA. JCSDA, UWisc, NASA).

Comparison between different techniques is an important task in planning for future operational systems. Our development of a unified, variational/ensemble-based WRF data assimilation system is key to ensuring clean comparisons. Given the complexity of real-world data assimilation systems, great care is needed to assess the strengths and weaknesses of the different approaches. MMM has strong expertise in both variational and ensemble-based data assimilation techniques, and so is well placed to perform detailed research comparisons. We will develop an objective assessment of the relative skill of 4D-Var and EnKF approaches using both Observing System Simulation Experiments (OSSEs) and real-world assimilations.

OSSEs make use of a defined "truth trajectory" to study data assimilation and forecast performance in an idealized setting. Observations may be simulated with error characteristics appropriate for existing, or proposed components of the observing system. OSSEs will provide a clean and flexible way to assess the potential impact of changes to assimilation technique and/or observation network. However, OSSE-type experiments do have weaknesses. In reality, neither observations nor forecast models obey the simplified error characteristics typically applied in OSSEs. It is therefore vital that we supplement OSSEs with real-world simulations to provide a more realistic assessment of the likely impact of changes to the initial conditions of the forecast (the analysis). To ensure statistically significant results, and to reduce the impact of model spin-up problems, real-world data assimilation experiments should be performed over extended trial periods. Extended evaluation periods will therefore be chosen representing classic cases of severe convective events, and landfalling hurricane forecasts.

Next section: Improving Forecasting of Pollutants and Hazardous Materials

 

 

© 2006, UCAR | Privacy Policy | Terms of Use | Webmaster