To forecast monsoon with Dynamical Model, supercomputer to be used by IMD
Science and Technology
India
The Union Ministry of Earth Sciences (MoES) in the last week of July 2016 announced that India Meteorological Department (IMD) will use supercomputer to forecast India’s annual summer monsoon.
The forecast made by a supercomputer will be based on a dynamical monsoon model. It will be operational from 2017.
What is Dynamical Monsoon Model?
• The dynamical model is also known as the Coupled Forecast System.
• It collates data on local as well as global weather patterns to simulate a forecast for a specific duration.
• The Indian Institute of Tropical Meteorology (IITM), Pune, has been using this model on an experimental basis for around 10 years to assist IMD’s annual monsoon forecasts.
• To operationalise the alternate model, IMD is investing in supercomputers that can function at a speed of 10 petaflops-per second system.
• The dynamical model has achieved a 60% accuracy at present and IMD aims to take it up to 77%.
• The dynamical model provides a wider scope to upgrade forecasts for specific regions, in sync with the changing weather patterns and variations of a progressing monsoon.
• The dynamical model and faster computing will also improve short range forecasts as the IMD will be able to access sharper resolution images.
Current Model to Predict Monsoon
• IMD has been using the ensemble statistical model to predict monsoon since 2007.
• A basic statistical model was in use first since 1920.
• In 2007, it switched to the ensemble statistical forecasting due to the inaccuracy of the annual summer forecasts.
• The existing model relies on arriving at a prediction based on historical monsoon data coupled with data on sea-surface temperatures and winds.
