Global Weather Assimilation and Modeling
- Intro to Earth System Modeling, FIM
- Icosahedral Grid in FIM, NIM
- FIM Real-Data Tests
- Global Observations for Assimilation, NCEP Gridpoint Statistical Interpolation
- Global Assimilation with Ensemble Kalman Filter
- Panel Discussion
Goal and Scope
The goal of global weather assimilation and modeling is to improve medium range weather forecasts, and increasingly, even short-range prediction. Global weather modeling is the key to NOAA's mission of predicting weather from 1 to 14 days in the future. Much of the associated public service, such as hurricane, winter storm and other economically important forecasts are dependent on the global weather models. This topic includes:
- Global scale assimilation - Use of in situ and satellite data, along with techniques such as variational analysis and Kalman filtering to determine the initial state of the global atmosphere.
- Development, implementation and improvement of the Earth System Modeling Framework.
- Dynamic Core - Development of new dynamic cores, particularly the hybrid coordinate core, for global models.
- Physics packages - Development of new physics packages for the global model.
- Ocean models - Ocean models are required for medium range weather prediction and for coupled atmospheric-ocean models.
- Studies - Use of global weather models to understand weather and climate. For example the use of the model ensemble for Dynamic Extended Range Forecasting.
- Global tactical forecasting - High-frequency data assimilation and modeling will extend to global coverage as satellite data assimilation techniques improve and global observations become available more quickly.
Rationale and Payoffs
Global weather models are crucial to improved medium range prediction. They are also a key part of determining the value of the global observing system because they are the measure, through Observing System Simulation Experiments (OSSEs) and Observing System Tests (OSTs), of the value of subsystems of the global observing system. Global model OSSE nature runs are also necessary prerequisites for performing regional-scale OSSEs.
There are several payoffs associated with global weather prediction. First, improved operational weather prediction is of great economic value to the nation. Transportation, industry and the public are able to use 1 to 14 day forecasts to economic advantage, particularly as these predictions become more reliable. Another payoff is in the improved understanding of the global system needed for climate modeling. A third important value of global modeling is the improved design of the global observing system based on systematic tests of observing system components. Observing system simulation experiments (OSSEs) are an important part of these design activities. Since a large fraction of NOAA's budget goes to the global observing system, this could be very important for optimum use of funds.
Major Collaborators and Their Research Foci
- Earth System Research Laboratory
- Global Systems Division: Existing expertise is in regional modeling, data assimilation, and OSSEs, but it has been developing concepts for global assimilation and modeling. Its expertise is in variational analysis, dynamic cores, and in physics packages, particularly moist physics.
- Physical Sciences Division: Extensive experience with the NCEP spectral model (Global Forecast System). They have been running these models on the ESRL supercomputer.
- Other NOAA
- NCEP EMC: Main repository of knowledge of global weather prediction in the U.S. They have extensive knowledge on global data assimilation, dynamic core, physics packages, and ocean models.
- Joint Center for Satellite Data Assimilation: Relatively new group, closely connected to NCEP, that is concentrating in techniques for development of radiative transfer models, and assimilation of all types of satellite data.
- Geophysical Fluid Dynamics Lab: The premier NOAA lab for global climate modeling, GFDL has decreased its interest in global weather modeling during recent years.
- NESDIS: Global satellite observations and data sets for model evaluation and diagnostic improvement.
- CIRA: Has extensive experience in global and regional modeling.
- NASA: Global observations and data sets for model evaluation and diagnostic improvement.
Contributions to NOAA Goals
- NOAA's Strategic Plan FY 2005-FY2010. Performance objective: Increase lead-time and accuracy for weather and water warnings and forecasts. Outcome: Reduced loss of life, injury and damage to the economy. (p. 9)
- Research in NOAA - A Five-Year Plan: Fiscal Years 2005 - 2009. "Community modeling approaches . . . (such as) the Weather Research and Forecast model . . .will accelerate the transfer of new models into operations." (Note: One of the efforts in global modeling will be to test the applicability of the WRF model to the global domain.) (p. 8)
- Science and Technology Infusion Plan 2004. Vision: Winter storms lead time increased from 9 hours to days. Hurricane warning lead-time from less than 24 hours to 3 days. Marine warnings increased to 3 days for gales and storms. (p. 8, 9)
Major Information Products, Customers, and Linkages
The direct customer for model assimilation and prediction development is NCEP EMC. There are a wide variety of customers for information from global models. For example, ESRL's Physical Sciences Division and other organizations have investigated the use of global weather models for Dynamic Extended Range Forecasting.