High Resolution Numerical Simulation of Philippine Weather and Climate using the Agila Beowulf Cluster

Previous simulations of regional scale weather and climate required the power of mainframe supercomputers to handle the formidable nonlinear fluid dynamical equations that describe the atmosphere. Using a reasonably powerful Linux-based Beowulf cluster (christened the Agila), we report a successful application of the Mesoscale Model 5 (MM5) on the Philippine domain. A high-resolution profile of the atmosphere was created for the month of February 2001 using a nested technique. This result was then validated with actual station observations and satellite pictures. In particular, we show good agreement of the model results with actual rainfall data and the observed path of a cyclone that passed through the country at this time. This preliminary success shows how it is now possible to more accurately characterize Philippine weather and climate even with meager computing resources.

[Loyola Schools Review, 2004]