Sunrise with stormy sky

End-to-End Modeling Systems

Developing and tailoring modeling systems for international partners optimized for the end-user

As high-resolution regional models (e.g., Weather Research and Forecasting: WRF) become more skillful, and computing hardware becomes less expensive, the use of models for meeting the specialized operational needs of various stakeholders is growing rapidly. While modeling software and hardware are easy to acquire, users are still faced with the following challenges:

  • optimizing the overall system performance to meet specific needs;
  • verifying the model skill from point to area scales;
  • developing specialized model-output products and visualization tools; and
  • providing training to users on enhancing modeling systems and using the high-resolution products.

Building upon expertise in numerical weather prediction (NWP) solutions, RAL has built a program developing and tailoring modeling systems for international partners optimized for the end-users’ region of interest and application.

Forecast verification and evaluation is often an unseen partner in engineering and scientific advances.  Verification provides systematic and objective evaluation of the quality (or performance) of a forecast.  In turn, this process allows comparisons to be made between versions in the development of an advanced forecast system.  RAL has expertise in developing customized verification systems to meet the needs of the end-user that provides tools (e.g., types of operational diagnostic statistics, display system, etc.) to provide an assessment of tailored modeling systems. 

Example of an End-to-End Operational Forecast System

NCAR has supported the modernization of the NWP capability for Saudi Arabia Meteorological Service (General Authority for Meteorology and Environmental Protection: GAMEP) through the development of a Weather Research and Forecasting (WRF) and a real-time data assimilation system.  A component of the NWP system was the implementation of the NCAR/RAL Joint Numerical Testbed (JNT) Model Evaluation Tools (MET) for real-time model evaluation, development of a forecast display system, and extensive forecaster and NWP system training.  An operational dust forecast system based on WRF-Chem was installed at GAMEP.  All the operational forecasts, dust forecasts, and verification products are available to GAMEP forecasters using customized web-based forecast display systems and Unidata’s interactive Integrated Data Viewer (IDV) display system. 

RAL is supporting the implementation of the next generation modeling system for GAMEP.  The current development includes the implementation of improved WRF-based modeling framework that includes advanced data assimilation, model verification, and ensemble techniques.  The system also includes the latest advancement of the dust modeling system and mobile application visualization technology.  Extensive training will be provided to GAMEP scientists and engineers.

Future Development Activities

RAL continues to respond to the needs of stakeholders, developing, customizing and transferring technology to provide end-to-end forecast solutions. These partnerships range from US-based applications with government and private institutions to international weather agencies.  A significant component of this end-to-end forecast system development effort is to build capacity through workshops and hands-on training of local scientists and engineers.  RAL continues to move the technology forward and is developing end-to-end modeling solutions that utilize cloud-computing services.  This solution is useful for international met services that limited computing and infrastructure resources to run an operational regional forecast system but have the need for forecast capabilities to predict and warn high impact weather events.

Contact

Paul Kucera

Mgr, International Cap & Bus Dev

email