This monsoon, IMD to begin special forecast for rain-fed areas
IMD will also use a multi-model ensemble forecast (a combination of different models) for the first time to predict monsoon rains this year
This monsoon, India Meteorological Department (IMD) will issue a special seasonal forecast for rain-fed areas where there is no irrigation to support agriculture.
IMD will also use a multi-model ensemble forecast (a combination of different models) for the first time to predict monsoon rains this year.
IMD held a day long brainstorming session on Monday to analyse the features of monsoon in 2020 and forecast plans for 2021.
“We were using some models like the Monsoon Mission Coupled Forecasting System (MMCFS) which is a dynamic model. Now we may have to use a combination of models to issue probabilistic forecasts and more accurate monsoon forecast. As an experiment we will also try issuing specialised forecasts for rain-fed areas this year,” said M Mohapatra, director general IMD.
IMD scientists said monsoon in 2020 was unique with large month-on-month variation and unusual rainfall patterns in four sub-regions of the country. An assessment of the seasonal forecast performance of IMD for monsoon 2020 was not found to be up to the mark. The error limit was breached for certain zones in the regional distribution forecast of IMD.
IMD had forecast normal monsoon rain in 2020 at 102% of long period average (LPA) with error margin of +/-4% for the season. But actual rainfall was above normal at 108.7% of LPA. IMD in its zonal monsoon forecast last year predicted 107% of LPA (with model error of +/-8%) monsoon rain over northwest India but the region recorded 84% of LPA; for northeast India it was predicted to be 96% of LPA but recorded 106%; for south peninsula it was predicted 102% but recorded 130% of LPA.
To address concerns of inaccurate forecasting of monsoon patterns in some years, IMD has decided to use a multi-model ensemble forecast.
Mohapatra said the reason monsoon 2020 was unique and difficult to forecast is because of intra-seasonal variability and formation of several low pressure systems which couldn’t be captured. Deficient rain or excess rain spells also couldn’t be predicted well. “There is need therefore to quantify the uncertainty in the monsoon forecast so that forecast can be used for various applications like energy, health, agriculture etc,” Mohapatra added.
IMD is also likely to revisit the monsoon withdrawal criteria this year. For the past ten years, withdrawal of monsoon has been significantly delayed compared to the normal withdrawal dates. IMD scientists will assess whether this is because of IMD’s strict monsoon withdrawal criteria, like no rain in the region for at least five days, establishment of anticyclonic wind pattern, considerable reduction in moisture content, or is it because monsoon withdrawal has actually delayed significantly in the past decade.
IMD’s modelling groups have been asked to improve their forecasts for river catchments as a new feature of providing five-day forecast (instead of three day) for sub-river catchments in the country is likely to be started this year. New forecasting models will also be deployed for forecasting of urban flooding in Chennai and Mumbai.
DS Pai, senior scientist, IMD Pune, said operational forecast error was highest between 1998 and 2002—over 10% with 1998 recording a forecast error of 21% and 2002 recording an error of 22%. But forecast error has improved in the past 13 (2008 to 2020) years to an average of 5.82% after an ensemble model was implemented. “But we need further improvement in models. IMD is now planning to use both Monsoon Mission Coupled Forecasting System (MMCFS) and National Centre for Medium Range Weather Forecasting (NCRMWF)’s models this year to generate monsoon forecast for better accuracy.”
Pai also highlighted that in a few years, India may be entering an epoch of above average monsoon rains. From 1945 to 1985 was also an above average rain epoch but from 1985 onwards, it has been a below average rain epoch.
“In a few years, we are likely to transition to an epoch of above average monsoon rains because monsoon follows epochs of around 31 years in these transitions. An above average epoch will be good for the country and agriculture but would mean more extreme rainfall events for which we should prepare,” added Mohapatra.
Monsoon rains are a lifeline for about 60% of the country’s net cultivated area, which have no irrigation.
The monsoon impacts inflation, jobs and industrial demand. Good farm output keeps a lid on food inflation. Ample harvests raise rural incomes and spending, helping inject demand into the economy.