Weather forecasts: India
(→Methods of forecast) |
(→2019) |
||
(4 intermediate revisions by one user not shown) | |||
Line 9: | Line 9: | ||
[[Category:S&T |W ]] | [[Category:S&T |W ]] | ||
[[Category:Climate/Meteorology |W ]] | [[Category:Climate/Meteorology |W ]] | ||
+ | |||
+ | =Discrepancies= | ||
+ | == IMD’s monsoon prediction vs. Skymet’s== | ||
+ | [https://epaper.timesgroup.com/Olive/ODN/TimesOfIndia/shared/ShowArticle.aspx?doc=TOIDEL%2F2018%2F10%2F01&entity=Ar00302&sk=DB7C233F&mode=text Amit Bhattacharya, IMD’s monsoon prediction off target, private Skymet update more accurate, October 1, 2018: ''The Times of India''] | ||
+ | |||
+ | |||
+ | The India Meteorological Department had forecast a normal monsoon in 2018, pegging the countrywide rainfall at 97% of the long period average. The season ended on Sunday at a rain deficit of 9.4%, making the national weather agency’s forecast off by more than the 4% margin of error. | ||
+ | |||
+ | While IMD’s monsoon forecasts have in recent years been getting more accurate, this year the department has not just been more than 6% off target, it’s categorisation of “normal” seasonal rainfall too proved inaccurate. The monsoon ended in the “below normal” zone, close to the deficient category (below 10%). | ||
+ | |||
+ | “It was the high monsoon shortfall in September that led to the inaccuracy. We were expecting 90-91% rainfall in September after taking into consideration an evolving El Nino. But high convection activity in northwest Pacific during the month drew moisture away from our region,” said D Sivananda Pai, IMD’s lead monsoon forecaster. | ||
+ | |||
+ | In contrast, private forecaster Skymet, which had initially predicted 100% monsoon rainfall this year, made a course correction midway into the season. | ||
+ | |||
+ | In an update released on August 1, Skymet downgraded the monsoon to “below normal” at 92% of the long period average — a forecast that came close to the actual figure of about 90.6%. | ||
+ | |||
+ | ===2018: IMD’s overall forecast vs. IMD’s regional predictions=== | ||
+ | [https://epaper.timesgroup.com/Olive/ODN/TimesOfIndia/shared/ShowArticle.aspx?doc=TOIDEL%2F2018%2F10%2F01&entity=Ar01605&sk=550DDEB9&mode=text Amit Bhattacharya, IMD’s regional rain forecasts prove to be more accurate, October 1, 2018: ''The Times of India''] | ||
+ | |||
+ | |||
+ | While IMD’s overall monsoon forecast missed the mark this year, the department’s regional rainfall predictions proved to be far more accurate. | ||
+ | |||
+ | The department had forecast good rainfall over northwest India at 100% of LPA. The region saw its best monsoon in five years which ended at 98% of LPA. Likewise, its forecasts for central India (99% versus actual of 94%) and south peninsula (95% versus 99%) were well within the 8% error margin. | ||
+ | |||
+ | East and northeast India, however, performed below expectations with just 76% rainfall during the season as against a forecast of 93%. | ||
+ | |||
+ | In its first two forecasts (in April and May), IMD had said the monsoon was likely to be 97% of LPA. In an update for the second half of the season (Aug-Sept) released on August 3, IMD had forecast 95% rainfall for the two months. The actual rainfall during the period was close to 87%. September itself ended with a massive shortfall of more than 23% and was the driest monsoon for the month since 2015. | ||
+ | |||
+ | In contrast, private forecaster Skymet had forecast poor rainfall in August (88%) and a slightly better performance in September (93%). The opposite turned out to be true, with August rainfall being slightly less than 93% of LPA. | ||
+ | |||
+ | In the event, fairly good rainfall distribution, despite the overall shortfall, kept the rain deficiency down to 12 subdivisions of the country out of 36 (33% of total). The deficiency was higher if one looked at the districts with 38% of districts for which IMD had data (252 out of 659) showing deficient or large deficient rainfall. | ||
+ | |||
+ | Last year, too, IMD had forecast a normal monsoon but it had ended up in the “below normal” range. The season’s deficit last year was 5%. However, the deviation of actuall rainfall from the forecast was within IMD’s 4% margin of error. | ||
=Methods of forecast= | =Methods of forecast= | ||
Line 52: | Line 85: | ||
==2009-16== | ==2009-16== | ||
− | |||
[[File: Forecasts vs actual monsoon, 2009-16.jpg|Forecasts vs actual monsoon, 2009-16; [http://epaperbeta.timesofindia.com/Gallery.aspx?id=19_04_2017_029_012_015&type=P&artUrl=Positive-signal-for-economy-as-IMD-predicts-normal-19042017029012&eid=31808 The Times of India], April 19, 2017|frame|500px]] | [[File: Forecasts vs actual monsoon, 2009-16.jpg|Forecasts vs actual monsoon, 2009-16; [http://epaperbeta.timesofindia.com/Gallery.aspx?id=19_04_2017_029_012_015&type=P&artUrl=Positive-signal-for-economy-as-IMD-predicts-normal-19042017029012&eid=31808 The Times of India], April 19, 2017|frame|500px]] | ||
− | See graphic | + | |
+ | '''See graphic''': | ||
+ | |||
+ | ''Weather forecasts: India, 2003-2015'' | ||
+ | |||
+ | == Arrival of monsoon, 2014-18== | ||
+ | [[File: Monsoon's onset date vs IMD's forecast, 2014-18.jpg|Monsoon's onset date vs IMD's forecast, 2014-18 <br/> From: [https://epaper.timesgroup.com/olive/ODN/TimesOfIndia/shared/ShowArticle.aspx?doc=TOIDEL%2F2019%2F05%2F16&entity=Ar00825&sk=D614BC08&mode=text Amit Bhattacharya, May 16, 2019: ''The Times of India'']|frame|500px]] | ||
+ | |||
+ | |||
+ | '''See graphic''': | ||
+ | |||
+ | '' Monsoon's onset date vs IMD's forecast, 2014-18 '' | ||
+ | |||
+ | ==2019== | ||
+ | [[File: Monsoon forecast vs actual, 2019.jpg|Monsoon forecast vs actual, 2019 <br/> From: [https://epaper.timesgroup.com/olive/ODN/TimesOfIndia/shared/ShowArticle.aspx?doc=TOIDEL%2F2019%2F10%2F01&entity=Ar01807&sk=205284FA&mode=text Amit Bhattacharya , Oct 1, 2019: ''The Times of India'']|frame|500px]] | ||
+ | |||
+ | '''See graphic''': | ||
+ | |||
+ | '' Monsoon forecast vs actual, 2019 '' | ||
+ | |||
+ | |||
+ | == 2011-2020: IMD got only two monsoon forecasts right== | ||
+ | [https://epaper.timesgroup.com/article-share?article=20_04_2022_001_017_cap_TOI April 20, 2022: ''The Times of India''] | ||
+ | |||
+ | [[File: June-September rain, 1st forecast vs actual, 2001-20..jpg|June-September rain, 1st forecast vs actual, 2001-20. <br/> From: [https://epaper.timesgroup.com/article-share?article=20_04_2022_020_004_cap_TOI Amit Bhattacharya, April 20, 2022: ''The Times of India'']|frame|500px]] | ||
+ | |||
+ | IMD has forecast a normal monsoon this year with rainfall pegged at 99% of long period average (LPA). How optimistic should you be of the projections coming true? Data reveals that in the last decade (2011-2020), just two of IMD’s first monsoon forecasts (made in April) were within 5 percentage points — the forecasts’ margin of error — of actual rainfall, reports Amit Bhattacharya. However, large deviations from actual were also down. Quantitative seasonal rainfall forecasts remain a huge scientific challenge and data shows IMD’s accuracy is actually improving. | ||
+ | |||
+ | [[Category:Climate/Meteorology|W WEATHER FORECASTS: INDIAWEATHER FORECASTS: INDIA | ||
+ | WEATHER FORECASTS: INDIA]] | ||
+ | [[Category:India|W WEATHER FORECASTS: INDIAWEATHER FORECASTS: INDIA | ||
+ | WEATHER FORECASTS: INDIA]] | ||
+ | [[Category:Pages with broken file links|WEATHER FORECASTS: INDIA | ||
+ | WEATHER FORECASTS: INDIA]] | ||
+ | [[Category:S&T|W WEATHER FORECASTS: INDIAWEATHER FORECASTS: INDIA | ||
+ | WEATHER FORECASTS: INDIA]] |
Latest revision as of 06:32, 6 May 2022
This is a collection of articles archived for the excellence of their content. |
Contents |
[edit] Discrepancies
[edit] IMD’s monsoon prediction vs. Skymet’s
The India Meteorological Department had forecast a normal monsoon in 2018, pegging the countrywide rainfall at 97% of the long period average. The season ended on Sunday at a rain deficit of 9.4%, making the national weather agency’s forecast off by more than the 4% margin of error.
While IMD’s monsoon forecasts have in recent years been getting more accurate, this year the department has not just been more than 6% off target, it’s categorisation of “normal” seasonal rainfall too proved inaccurate. The monsoon ended in the “below normal” zone, close to the deficient category (below 10%).
“It was the high monsoon shortfall in September that led to the inaccuracy. We were expecting 90-91% rainfall in September after taking into consideration an evolving El Nino. But high convection activity in northwest Pacific during the month drew moisture away from our region,” said D Sivananda Pai, IMD’s lead monsoon forecaster.
In contrast, private forecaster Skymet, which had initially predicted 100% monsoon rainfall this year, made a course correction midway into the season.
In an update released on August 1, Skymet downgraded the monsoon to “below normal” at 92% of the long period average — a forecast that came close to the actual figure of about 90.6%.
[edit] 2018: IMD’s overall forecast vs. IMD’s regional predictions
While IMD’s overall monsoon forecast missed the mark this year, the department’s regional rainfall predictions proved to be far more accurate.
The department had forecast good rainfall over northwest India at 100% of LPA. The region saw its best monsoon in five years which ended at 98% of LPA. Likewise, its forecasts for central India (99% versus actual of 94%) and south peninsula (95% versus 99%) were well within the 8% error margin.
East and northeast India, however, performed below expectations with just 76% rainfall during the season as against a forecast of 93%.
In its first two forecasts (in April and May), IMD had said the monsoon was likely to be 97% of LPA. In an update for the second half of the season (Aug-Sept) released on August 3, IMD had forecast 95% rainfall for the two months. The actual rainfall during the period was close to 87%. September itself ended with a massive shortfall of more than 23% and was the driest monsoon for the month since 2015.
In contrast, private forecaster Skymet had forecast poor rainfall in August (88%) and a slightly better performance in September (93%). The opposite turned out to be true, with August rainfall being slightly less than 93% of LPA.
In the event, fairly good rainfall distribution, despite the overall shortfall, kept the rain deficiency down to 12 subdivisions of the country out of 36 (33% of total). The deficiency was higher if one looked at the districts with 38% of districts for which IMD had data (252 out of 659) showing deficient or large deficient rainfall.
Last year, too, IMD had forecast a normal monsoon but it had ended up in the “below normal” range. The season’s deficit last year was 5%. However, the deviation of actuall rainfall from the forecast was within IMD’s 4% margin of error.
[edit] Methods of forecast
[edit] 2017/ a change to dynamic model
Jacob Koshy, Supercomputer to forecast monsoon with dynamical model, July 27, 2016: The Hindu
A dynamical monsoon model works by simulating the weather on powerful computers and extrapolating it over particular timeframes, which will be used by the IMD from 2017.
Though this method is normally effective in forecasting weather over a few days, using it to forecast the annual monsoon over 3 or 4 months has proved difficult.
While such models have been used for research purposes for long, the India Meteorological Department (IMD) has never integrated them into its operational forecast.
Differences between the traditional and dynamic model :
Traditional approach:
• An ensemble model, a statistical technique that uses an average of 6 meteorological values correlated to the monsoon such as sea surface temperatures in the Pacific, and North Atlantic sea level pressure. These values are derived from century-old meteorological data linked to the historical performance of the monsoon.
• Has failed to predict monsoon failures, in 2002 and 2004, leading to calls by meteorologists for a new, modern system.
Dynamic model:
• Called the Coupled Forecast System version 2.
• Has far achieved only 60% accuracy in forecasting the monsoon.
• A confidence boost came when the dynamical model's forecast along with the traditional one, its plans to give prominence to the dynamical model signals a new approach.
• This is a precursor to giving monsoon predictions over India’s 36 sub-divisions rather than only four broad geographic regions that encompass them. A dynamical approach can also be more easily tuned to account for rapidly changing global weather conditions.
[edit] Monsoon, actual and forecast
[edit] 1960-2013
See graphic, ' Monsoon, actual and forecast, 1960-2013 '
[edit] 2000-15
![](/ind/images/1/1b/Monsoon%2C_actual_and_forecast%2C_2000-15.jpg)
See graphic, ' Monsoon, actual and forecast, 2000-15 '
[edit] 2003-2015
![](/ind/images/4/41/Weather_forecasts_india.jpg)
See graphic, 'Weather forecasts: India, 2003-2015'
[edit] 2009-16
![](/ind/images/6/6e/Forecasts_vs_actual_monsoon%2C_2009-16.jpg)
See graphic:
Weather forecasts: India, 2003-2015
[edit] Arrival of monsoon, 2014-18
![](/ind/images/d/d3/Monsoon%27s_onset_date_vs_IMD%27s_forecast%2C_2014-18.jpg)
From: Amit Bhattacharya, May 16, 2019: The Times of India
See graphic:
Monsoon's onset date vs IMD's forecast, 2014-18
[edit] 2019
![](/ind/images/2/2e/Monsoon_forecast_vs_actual%2C_2019.jpg)
From: Amit Bhattacharya , Oct 1, 2019: The Times of India
See graphic:
Monsoon forecast vs actual, 2019
[edit] 2011-2020: IMD got only two monsoon forecasts right
April 20, 2022: The Times of India
![](/ind/images/b/b9/June-September_rain%2C_1st_forecast_vs_actual%2C_2001-20..jpg)
From: Amit Bhattacharya, April 20, 2022: The Times of India
IMD has forecast a normal monsoon this year with rainfall pegged at 99% of long period average (LPA). How optimistic should you be of the projections coming true? Data reveals that in the last decade (2011-2020), just two of IMD’s first monsoon forecasts (made in April) were within 5 percentage points — the forecasts’ margin of error — of actual rainfall, reports Amit Bhattacharya. However, large deviations from actual were also down. Quantitative seasonal rainfall forecasts remain a huge scientific challenge and data shows IMD’s accuracy is actually improving.