SEA Seminar - Building Python-Based Operational Systems for Prediction of Atmospheric Processes

In the context of atmospheric models, operational systems are those that dependably provide important, time-critical forecasts for use in decision-support systems.  They are characterized by the need for rapid data acquisition and model setup, reliable execution of the model, and post-processing activities that sometimes require delivery of model output timesteps as soon as they are completed.  In Alaska, reliable and timely weather forecasts are essential for a variety of commercial, recreational, and simple day-to-day living activities.  In addition to routine operational forecasting, Alaskans need to be prepared for surprise events.  Volcanic eruptions occur with some regularity in Alaska and Russia, affecting air transport and sometimes shutting down urban areas.  Uncontrolled wildfires in unpopulated regions are a fact of life, and Alaskans are sometimes subject to huge smoke emissions during the summer.   Additionally, Alaska is affected by non-predictable anthropogenic events such as toxic emissions and nuclear fallout.

 In addition to providing the tools for decision support, operational products from Alaska are also useful in a number of "downstream" research endeavors including road weather prediction systems, aviation icing products and greenhouse gas studies.  

 With a decade of experience in provision of operational products, often for large areas at high resolutions, we have gained a fair amount of experience and wisdom, and still have much to learn.  In this presentation we discuss some of the systems we have deployed, lessons learned, and the many exciting ambitions we have.


Room Number: 
1022 - Large Auditorium

Type of event:

Will this event be webcast to the public by NCAR|UCAR?: 
Calendar Timing: 
Thursday, June 20, 2013 - 3:00pm