Mesoscale Forecast Model (ARPS)
Select Current Run Loop:
Select Previous Run Loop:
| Note: The mesoscale prognostics depicted on these pages are experimental and may not be available at all times. The accuracy or reliability of the data is not guaranteed nor warranted in any way. The data is provided as is and should not be used as the sole resource for decision making. Mesoscale forecast guidance is best interpreted by a professional meteorologist who is familiar with the particular modeling system, including any model biases. This version of the ARPS model is greatly influenced by the ADAS analyses, the data sets in those analyses, and the quality of the RUC model that provides the boundary conditions. |
Model overview: The ARPS mesoscale model is available from the Center for Analysis and Prediction of Storms in Norman Oklahoma. The Melbourne National Weather Service (MLB) version of the model runs on a 10 node, 20 processor Linux cluster. The current hardware allows for the generation of a 0 to 9 hour forecast within 3 hours. The ARPS output is on the same 177 by 177 domain and 4-km horizontal resolution as ADAS. The model contains 45 vertical levels and current post-processing strategies produce a temporal resolution of 30 minutes. Model output is post-processed as forecast increments are finished, providing guidance to the forecaster as timely as possible. The 9 hour forecast length was chosen to maximize the frequency of forecast updates with fresh ADAS observational analyses under the current hardware constraints. Four ARPS runs are initialized each day at 3, 9, 15 and 21 UTC. The MLB ARPS uses the 3-12 hour, 40-km RUC forecast as its background field and lateral boundary conditions. At the start of each model run, an Intermittant Data Assimilation (IDA) cycle is completed. The IDA cycle incorporates observational data into the ARPS model by alternating between ADAS analyses that initialize very short range ARPS forecasts over a specified time interval (Case 2003). The IDA is typically run over a one hour of pre-forecast, but can be adjusted to any length of time. An IDA strategy has been instituted to start the forecast model with as much mesoscale detail as possible. At the start of the IDA the RUC 2-h forecast is used as the first guess field for the initial ADAS. This first ADAS analysis is then used to initialize a short 15 minute ARPS forecast interval. Following this 15 minute forecast interval a second ADAS analysis is generated with a complete set of new observational data and using the ARPS 15 minute forecast as its first guess field. This process of generating new ADAS analyses and then 15 minute ARPS forecasts continues through the entire IDA cycle. The last ADAS analysis is used to initialize ARPS as it begins the 9 hour forecast. For more information about the Mesoscale Forecast Model (ARPS) or ADAS contact Pete Blottman or Scott Spratt
If you have questions about local modeling efforts at WFO Melbourne contact Dave Sharp
|
