In 1987, when India was hit by a spell of massive drought, the then Prime Minister of India, Rajiv Gandhi summoned his scientists at the weather department and asked like any of the farmers in the field would have asked, a simple question, “What has happened?”. The reply was equally simple, “Monsoon failure.”
“I don't want to hear that answer again.” Rajiv Gandhi retorted. Subsequently, he created a range of institutions to predict weather phenomena. Less than a third of the country had received normal rainfall that year and predictions and hopes lost meaning with the hardest-hit areas receiving as low as 75 percent below normal rainfall.
When GSHLV Prasada Rao narrated this episode, India is in a dejavu of drought. Will 2012 be another 1987?
It may or may not be.
GSLHV Prasada Rao or Gondi Sri Lakshmi Hari Vara Prasada Rao Ph.D is the Special Officer, Academy of Climate Change Education and Research (ACCER) at Kerala in South India.
He explains that science part of Monsoon prediction is not understood in India, as it should be.
“First of all one should understand that no where else in the world, this sort of prediction is being practiced. So the task in itself is unique and challenging.”he said.
“Another important factor is that there are two types of predictions in this arena: weather prediction and climate prediction and two are different.
Monsoon prediction is about weather prediction and the latter is composed of four types:
1. Short range weather forecast
2. Medium range weather forecast
3. Long range weather forecast
4. Nowcasting which is employed in local conditions and predicted3-4 hours before.
2. Medium range weather forecast
3. Long range weather forecast
4. Nowcasting which is employed in local conditions and predicted3-4 hours before.
Monsoon prediction is about long range weather forecasting.” Prasada Rao added.
When asked about the difficulty aspect of predicting monsoons, he brought attention to historical facts.
Monsoon prediction is about statistical modelling. Earlier, predictions were doled out for a period ranging between 3 days and 10 days. This was subsequent to the drought episode of 1987; IMD or Indian Meteorological Department revived the British practice of giving out monsoon predictions in 1988.
In 2002, this model of monsoon prediction failed. Droughts again hit India. Then IMD was employing ARIMA(Auto Regressive Integrated Moving Average) Model constituted by 16 parameters.
The 2002 event of drought forced IMD to revise its methods and it then employed an ARIMA model with eight parameters. The new model predicted monsoons 45 days in advance of June 1.
Also, in May, 8/10 ARIMA model was utilised to alert farmers. For 2003, this arrangement worked out well.
But in 2004, the model failed and in 2006, a 5/6 ARIMA model was employed. This too worked out well until 2009,when it again failed.
“Since 2010, if I am right, new models have been introduced.” he said.
So what ails India in monsoon prediction?
It has to be taken note that the science part of monsoon prediction is not well understood by us. Interaction between land features, ocean and atmosphere is not well deciphered. The Physics part of it is still obscure. And when it comes to scientific and technical side, we also need right persons in the right chairs, Rao said.
India's climate prediction infrastructure is excellent. There is enough talent out there. But compared to scientists in the West, we lack commitment. We lack international collaboration with institutes of repute. More over, we should develop our own models;currently we subscribe to models—Global Climate Models-- from other nations and just extrapolate.
Public-Private partnership is also important. The government has been doing the business of prediction for the past 65 years! We should not leave that to government sector alone. Private players too should chip in.
Scientists outside India are nowhere superior to us in talent. But they have an element of 'commitment superiority', he concluded.
No comments:
Post a Comment