Initialization Improvements to a Quasi-Operational Local-Domain Mesoscale Model

John S. Snook Paul Schultz John McGinley

NOAA Forecast Systems Laboratory Boulder, CO 80303

1. Introduction

A local-area data assimilation system, called the Local Analysis and Prediction System (LAPS), has been developed at the National Oceanic and Atmospheric Administration's Forecast Systems Laboratory (FSL). LAPS merges all relevant available data sources into surface and upper-air isobaric analyses (Albers et al. 1996, McGinley 1995). The analyses are used to initialize mesoscale numerical prediction models to generate high-resolution, local-domain forecasts (Snook and Pielke 1995, Snook et al. 1995). The complete LAPS system is designed to operate in real time within a local forecast office, where local datasets are readily available and LAPS products are efficiently posted to the operational forecaster. It has been demonstrated that LAPS analyses and predictions can provide additional useful guidance toward the support of operational mesoscale forecasts (Snook et al. 1998, Snook et al. 1997, Stamus and McGinley 1997). Model verification results suggest a close correlation between the quality of the LAPS initialization and the model forecast performance. This paper discusses the latest efforts to improve the LAPS initialization used by the mesoscale prediction models.

2. Quasi-operational LAPS

The analysis portion of LAPS has been running quasi-operationally at FSL since 1989. Analyses and derived products are available to the FSL prototype Advanced Weather Interactive Processing System (AWIPS) meteorological workstation (MacDonald and Wakefield 1996) that is used for an in-house daily weather briefing, to the National Weather Service Forecast Office (NWSFO) in Denver, and to the LAPS World Wide Web (WWW) home page (http://laps.fsl.noaa.gov). Feedback from all of these sources is incorporated into ongoing developments and improvements to LAPS.

The predictive component of LAPS is designed to utilize any available nonhydrostatic mesoscale forecast model. The Regional Atmospheric Modeling System (RAMS) (Pielke et al. 1992, Walko et al. 1995) developed at Colorado State University was incorporated into the quasi-operational LAPS in 1992 and the Penn State University/National Center for Atmospheric Research fifth-generation Mesoscale Model} (MM5) (Grell et al. 1995) was included during early 1996. Both models, initialized with identical LAPS analyses, have been producing 12-hour forecasts twice per day for nearly two years. Details of the mesoscale model configuration are discussed by Snook et al. (1998), Snook and Pielke (1995), and Snook et al. (1995). Model output is available to the FSL workstation and to the LAPS website home page.

Other LAPS demonstrations have included the Seattle NWSFO, the University of Oklahoma, the Chinese Meteorological Administration, and the first operational implementation at the Peachtree City NWSFO in support of the 1996 Centennial Olympic Games (Snook et al. 1998, Snook et al. 1997, Stamus and McGinley 1997). All of these venues successfully demonstrated the utility of LAPS as a useful operational mesoscale forecasting tool.

3. Improvements to model initialization

Generating quasi-operational local-area model output simultaneously from two separate mesoscale forecast models has produced a unique dataset. Figures 1a, 1b, and 1c illustrate a comparison of MM5 and RAMS forecast verification for all predictions generated from 14 March 1996 through 21 September 1997. Statistics are computed by a direct comparison of all available surface observations with model output interpolated to the observation location. Although RAMS and MM5 are substantially different forecast models, the verification results are similar. Subjective evaluation of the predictions on a daily basis supports the objective results. Typically, when one model does well, they both do well, and similarly, when one model does poorly, they both perform poorly. These results suggest the significance of the LAPS initialization and the forecast lateral boundary conditions used by the models to the forecast performance. In a separate study, the effects of the forecast lateral boundary conditions are being investigated through parallel model forecasts using a nested grid configuration that essentially displaces the effects of the forecast lateral boundary condition farther away from the area of interest.

Improvements to the LAPS analyses have typically comprised adjustments and technique improvements to individual analyses. Recently, a more concerted effort has been initiated to investigate initialization improvements that might be realized through four-dimensional data assimilation (4DDA), which potentially combines the strengths of both components of LAPS, analysis and predictive, to provide improved analysis and forecast products.

A case study investigation (Cram et al. 1993) using 4DDA techniques within LAPS has suggested that improvements can be realized. Two techniques were attempted. One used a nudging approach where the model fields were nudged toward LAPS analyses for three hours of model integration followed by 12 hours of forecast integration. The other approach took advantage of a reanalysis technique where forecast model output was used as an improved background field for the generation of future LAPS analyses. These LAPS analyses were then used to initialize the next forecast model.

We are currently testing these 4DDA techniques in the quasi-operational LAPS system. Parallel forecasts will be generated to evaluate the differences. Results from these experiments will be presented at the conference.

4. Summary

LAPS is a complete local-area data assimilation system designed to function in the local forecast office. Quasi-operational analyses from LAPS have been used to initialize the RAMS and MM5 mesoscale forecast models twice per day. Verification of the model output with surface observations suggests that model performance is closely tied to the quality of model initialization. Hence, improvements to the LAPS initialization fields are an ongoing development. A current investigation focuses on the use of two types of 4DDA schemes that attempt to combine the strengths of the analysis and predictive components of LAPS to produce improved LAPS products. Results from this investigation will be presented at the conference.

5. Acknowledgements

The authors wish to thank Drs. Roger Pielke and William Cotton of Colorado State University and Dr. Craig Tremback of Mission Research Corporation for their continued permission to use RAMS for this project. Pete Stamus reviewed the article and Nita Fullerton provided technical editing support. RAMS was developed under the support of the National Science Foundation (NSF) and the Army Research Office (ARO).

6. References

Albers, S. C., J. A. McGinley, D. L. Birkenheuer, and J. R. Smart, 1996: The Local Analysis and Prediction System (LAPS): Analysis of Clouds, Precipitation, and Temperature. Wea. Forecasting, 11, 273-287.

Cram, J. M., J. S. Snook, S. C. Albers, and J. A. McGinley, 1993: A brief description of, and preliminary results from, the LAPS modeling/4dda system. Preprints, Fifth International Conference on Aviation and Weather Systems, Vienna, VA, Amer. Meteor. Soc., J10-J14.

Grell, G. A., J. Dudhia, and D. R. Stauffer, 1995: A description of the fifth-generation Penn State NCAR Mesoscale Model (MM5). NCAR Technical Note TN-398+STR. 122 pp.

MacDonald, A. E., and J. S. Wakefield, 1996: WFO-Advanced: An AWIPS-like prototype forecast workstation. Preprints, 12th International Conference on Interactive and Processing Systems (IIPS) for Meteorology, Oceanography, and Hydrology, Atlanta, GA, Amer. Meteor. Soc., 190-193.

McGinley, J. A., 1995: Opportunities for high resolution data analysis, prediction, and product dissemination within the local weather office. Preprints, 14th Conference on Weather Analysis and Forecasting, 15-20 January 1995, Dallas, TX, Amer. Meteor. Soc., 478-485.

Pielke, R. A., W. R. Cotton, R. L. Walko, C. J. Tremback, W. A. Lyons, L. D. Grasso, M. E. Nicholls, M. D. Moran, D. A. Wesley, T. J. Lee, and J. H. Copeland, 1992: A comprehensive meteorological modeling system - RAMS. Meteor. Atmos. Phys., 49, 69-91.

Snook, J. S., P. A. Stamus, J. Edwards, Z. Christidis, and J. A. McGinley, 1998: Local-domain mesoscale analysis and forecast model support for the 1996 Centennial Olympic Games. Accepted Wea. Forecasting.

Snook, J. S., Z. Christidis, J. Edwards, and J. A. McGinley, 1997: Forecast results from the local-domain mesoscale model supporting the 1996 Summer Olympic Games. Preprints, 13th International Conference on Interactive and Processing Systems (IIPS) for Meteorology, Oceanography, and Hydrology, Long Beach, CA, Amer. Meteor. Soc., 26-30.

Snook, J. S., and R. A. Pielke, 1995: Diagnosing a Colorado heavy snow event with a nonhydrostatic mesoscale numerical model structured for operational use. Wea. Forecasting, 10, 261-285.

Snook, J. S., J. M. Cram, and J. M. Schmidt, 1995: LAPS/RAMS: A nonhydrostatic mesoscale numerical modeling system configured for operational use. Tellus, 47A, 864-875.

Stamus, P. A., and J. A. McGinley, 1997: The Local Analysis and Prediction System (LAPS): Providing weather support to the Olympic Games. Preprints, 13th International Conference on Interactive and Processing Systems (IIPS) for Meteorology, Oceanography, and Hydrology, Long Beach, CA, Amer. Meteor. Soc., 11-12.

Walko, R. L., W. R. Cotton, J. L. Harrington, and M. P. Meyers, 1995: New RAMS cloud microphysics parameterization. Part I: The single-moment scheme. Atmos. Research, 38, 29-62.