[ad_1]
NEW DELHI: How do the India Meteorological Department’s monsoon forecasts compare with actual rainfall in a year? In the last 10 years, IMD‘s first long range forecasts (LRF-1, issued in April) of the monsoon have on average been off the mark by 6.9%, which is higher than the forecasts’ stated error margin of 5%. The departure of the forecasts from actual rainfall was nearly the same in the preceding 10-period (2003-2012) at 7%. The department’s LRF-2, issued in May-June, were slightly more accurate, with an average departure of 5.9% in the last 10 years. Experts felt these departures weren’t too high.
This difference has remained largely constant over the past 20 years, with the average departure of the forecasts from actual monsoon rainfall during the 2003-2012 period being 7%.
What it means is that, more often than not, the actual rainfall deviates from the IMD forecasts by a value higher than the margin of error. However, given the complexities and uncertainties built into the exercise, many experts feel that these departures aren’t too high. “An average departure of 6.9% is not very large and is nothing to worry too much about. The actual numbers announced in the quantitative forecasts should be seen as indicative. The forecast is issued in April for periods that are up to five-six months away,” said Madhavan Nair Rajeevan, former secretary in the earth sciences ministry.
As can be expected, the department’s forecasts become more accurate closer to the monsoon season, which lasts from June to September. Often, the second forecast offers IMD a chance for course correction. In the last 10 years, the average deviation from actual rainfall of IMD’s second long range forecasts (LRF-2), issued in May/June, was 5.9%. However, these forecasts also had a narrower error margin of 4%.
In the last 10 years, the April forecast has come very close to the actual rainfall on two occasions – 2017 and 2021, when the difference was just 1%. In all other years, the predictions have been off by more than the 5% error margin. The predictions have been way off course in three of these years – 2019 when the difference was 14%, and 2016 and 2019 (each 9%). The complexities increase with forecasts for individual regions of the country, which is done during the second long range forecast.
“No other country in the world makes number-wise predictions in their seasonal forecasts, giving instead probabilistic outlooks for various possibilities. But in India, it has become a tradition which is now difficult to change,” Rajeevan said.
This difference has remained largely constant over the past 20 years, with the average departure of the forecasts from actual monsoon rainfall during the 2003-2012 period being 7%.
What it means is that, more often than not, the actual rainfall deviates from the IMD forecasts by a value higher than the margin of error. However, given the complexities and uncertainties built into the exercise, many experts feel that these departures aren’t too high. “An average departure of 6.9% is not very large and is nothing to worry too much about. The actual numbers announced in the quantitative forecasts should be seen as indicative. The forecast is issued in April for periods that are up to five-six months away,” said Madhavan Nair Rajeevan, former secretary in the earth sciences ministry.
As can be expected, the department’s forecasts become more accurate closer to the monsoon season, which lasts from June to September. Often, the second forecast offers IMD a chance for course correction. In the last 10 years, the average deviation from actual rainfall of IMD’s second long range forecasts (LRF-2), issued in May/June, was 5.9%. However, these forecasts also had a narrower error margin of 4%.
In the last 10 years, the April forecast has come very close to the actual rainfall on two occasions – 2017 and 2021, when the difference was just 1%. In all other years, the predictions have been off by more than the 5% error margin. The predictions have been way off course in three of these years – 2019 when the difference was 14%, and 2016 and 2019 (each 9%). The complexities increase with forecasts for individual regions of the country, which is done during the second long range forecast.
“No other country in the world makes number-wise predictions in their seasonal forecasts, giving instead probabilistic outlooks for various possibilities. But in India, it has become a tradition which is now difficult to change,” Rajeevan said.
#Imd #years #IMDs #1st #forecasts #India #News
