MUMBAI (TIP): India is finally ready with an indigenous model to predict long range and seasonal rainfall. In the next six months, the Indian Institute of Tropical Meteorology (IITM), Pune, will be able to operationalise a dynamic numerical climate model it has developed to forecast extended-range rainfall for a timescale of 15-20 days after results from the experimental stage were found to be reliable.

“For the first time, we have indigenously developed a prediction system that was not available in India. The tests have been done, the performance of the experimental model was found to be good and can be operationalised. We can now forecast rain every five days for 15-20 days especially for the agriculture sector which is the mainstay of the economy,” M Rajeevan, director, IITM  said.

India Meteorological Department (IMD), the official agency to disseminate weather forecast, will, however, validate the operationalisation of the model. At present, the IMD uses statistical model that needs lengthy calculations to make predictions of the southwest monsoon.

Scientists at IITM tasked with improving long range and seasonal scale forecasts (16 days to a season) have localised the indigenous model factoring Indian climatic conditions using the Climate Forecast System (CFS) developed in the US.

Developing the CFS-based Grand Ensemble Prediction System (CGEPS) is part of the National Monsoon Mission that was launched by the government in 2012 to improve accuracy of monsoon prediction especially since 60% agriculture in India is rain-fed.

Under the joint India-US collaboration, scientists used the National Centers for Environmental Predictions-Climate Forecast System (NCEP-CFSv2) model for monsoon prediction.

As part of the experimental version of the indigenous model, IITM has been providing forecasts for the agriculture sector to the IMD that issues advisories to the farmers over the last three years.

“The skillful and timely forecasts of intra-seasonal monsoon rainfall possess a greater potential utility for agriculture and water resource management. In the extended range, especially beyond the weather scale (two–three weeks), a single deterministic rainfall forecast is not sufficient thus the user community should also be given probabilistic forecasts that quantify the uncertainty,” said Atul Kumar Sahai, project director, National Monsoon Mission.

Sahai added, “The thumb rule is that before moving from an experimental version of any forecasting system towards operational version, its skill should be evaluated in detail.”

Suryachandra Rao, senior scientist at IITM, said after the monsoon mission, the model in the next phase will incorporate hydrological forecasting that will help in decisions relating to management of dam water -how much water can be released or whether the water can be used to generate hydropower.