Press Release

New Delhi, 30 June 2007

 

INDIA METEOROLOGICAL DEPARTMENT

 

Long Range Forecast Update

For 2007 South-West Monsoon Season Rainfall

 


1.Background

Since 2003, India Meteorological Department (IMD) has been following a two-stage forecast strategy for the southwest monsoon rainfall over the country as a whole, in which the first forecast is issued in April with the data up to March and the forecast update is issued by end of June with the data up to May. This year, IMD introduced the following new statistical forecast models for operational long range forecast of southwest monsoon rainfall over the country as a whole.

a)A new 5- parameter statistical forecasting system requiring data up to March, for the first forecast in April.
b)A new 6- parameter statistical forecasting system requiring data up to May for the forecast update in June.


The most important aspect of the new forecast system is the introduction of the concept of ensemble forecasts. For developing the models, two different statistical techniques namely, Multiple Regression (MR) and Projection Pursuit Regression (PPR) were considered.

Using the new 5-parameter statistical model, on19 April, IMD issued the following forecast for the 2007 southwest monsoon rainfall over the country as a whole.

"IMD's long range forecast for the 2007 south-west monsoon season (June to September) rainfall is that the rainfall over the country as a whole is likely to be 95% of the long period average with a model error of ± 5%."

2.Second Stage Forecasts

Now, with the availability of data up to May, IMD has prepared the second stage forecasts, which are being released now:

a)Forecast update for the 2007 southwest monsoon rainfall over the country as a whole using the new 6-parameter ensemble statistical forecast system. The forecasting system has an error of ± 4%. The Long period average of southwest monsoon rainfall averaged over the country as a whole is 89 cm (based on the 1941-1990 data) with a coefficient of variation of 10%.

b)Forecast for the rainfall over the country as a whole in the month of July based on a 6-parameter model, which has a model error of ± 9%. The Long period average of monsoon rainfall in the month of July is 293 mm (based on the 1941-1990 data) with a coefficient of variation of 13%.

c)Forecasts for the South-west Monsoon Season(June-September) rainfall for the following four broad geographical regions of India with a model error of ± 8%:

Northwest India - Jammu and Kashmir, Himachal Pradesh, Punjab, Rajasthan, Haryana, Chandigarh, Delhi, Uttaranchal and Uttar Pradesh.

Northeast India - Arunachal Pradesh, Meghalaya, Assam, Nagaland, Manipur, Mizoram, Tripura, Sikkim, West Bengal, Bihar and Jharkhand.

Central India - Gujarat State, Madhya Pradesh, Chattisgarh, Maharashtra, Goa and Orissa.

South Peninsula - Andhra Pradesh, Karnataka, Tamil Nadu, Kerala, Lakshadweep and Andaman and Nicobar Islands.

The long period average rainfall and coefficient of variation of rainfall of the 4 broad geographical regions of India are given below:

Area Long period average (mm) Coefficient of variation (%)
NW India
612
19
Central India
994
14
NE India
1429
8
South Peninsula
725
15


3.Experimental Forecasts

As a part of ongoing efforts to improve the long range forecast capabilities, experimental forecasts for the 2007 south-west monsoon rainfall based on the IMD's dynamical forecast system (Seasonal Forecast model of the Experimental Climate Prediction Centre (ECPC), USA) were also generated. Forecasts have been generated using observed sea surface temperature data of May and using the persistence method.

In addition, IMD has also taken into account the experimental forecasts prepared by national institutes like Indian Institute of Tropical Meteorology, Pune, National Centre for Medium Range Weather Forecast (NCMRWF), Noida, Indian Institute of Science, Bangalore, Space Applications Centre, Ahmedabad, National Aerospace Laboratory, Bangalore and Centre for Mathematical Modelling and Computer Simulation (CMMACS) Bangalore and operational/experimental forecasts prepared by international institutes like the National Centers for Environmental Prediction (NCEP), USA, International Research Institute for Climate and Society (IRI), USA, Meteorological Office, UK, the European Center for Medium Range Weather Forecasts(ECMWF), UK, the Tokyo Climate Centre and the Experimental Climate Prediction Center (ECPC), USA.

4.Onset and Advance of Monsoon 2007

Southwest monsoon advanced over south Andaman Sea on 11 May, almost 10 days earlier than the normal date. IMD predicted that the monsoon onset over Kerala this year would take place on 24th May with a model error of ±3 days. Monsoon advanced over Kerala on 28th May, 4 days earlier than the normal date of 1st June. By 29 May, monsoon covered the entire coastal Karnataka and some parts of south interior Karnataka and south Tamil Nadu. After this, there was hiatus in the advancement of monsoon for about one week. The hiatus was caused by the formation of the Super Cyclone "Gonu" over Arabian Sea, and its northwestward movement towards the Arabian coast.

On 8 June, monsoon advanced over some more parts of south interior Karnataka, Tamil Nadu, the entire Nagaland-Manipur-Mizoram-Tripura, some parts of Assam & Meghalaya and Arunachal Pradhesh. By 10 June, it covered most parts of Northeast India and advanced over some parts of Bihar. Between 13 to 18 June, monsoon covered Konkan and Goa, Andhra Pradesh, Orissa, West Bengal, most parts of Orissa and some parts of Maharashtra, Orissa, Chattisgargh and Jharkhand. On 26 June, southwest monsoon has further advanced into some parts of east Uttar Pradesh, remaining parts of Uttarakhand, entire Himachal Pradesh, Jammu and Kashmir and northern parts of west Uttar Pradesh, Haryana and Punjab. On 28 June, Southwest monsoon has further advanced into the remaining parts of Madhya Pradesh & Uttar Pradesh, most parts of East Rajasthan, some parts of west Rajasthan and some more parts of Haryana (including Delhi) and Punjab. As on 28 June, the northern limit of monsoon passes through Barmer, Jodhpur, Ajmer, Jaipur, Narnaul, Rohtak, Ludhiana and Amritsar.

The accumulated seasonal rainfall over the country as a whole during the period 1-27 June was 107 % of long period average.

5. La Nina Conditions over the equatorial Pacific

The warm sea surface temperature (SST) anomalies over the east equatorial Pacific associated with the moderate El Nino of 2006 have disappeared during February 2007. By the end of February 2007, SSTs were near average in the vicinity of the date line, and below average over the eastern equatorial Pacific. ENSO-neutral conditions continued in the tropical Pacific during May and June 2007. Equatorial SSTs during the last four weeks were more than 1ºC below average in areas between 120ºW and the South American coast, and more than 0.5oC above average in portions of the western Pacific. For the week ending 13 June, the SST anomalies were +0.2ºC and +0.4ºC in the Niño 3.4 and Niño 4 regions respectively.

Most statistical models show ENSO-neutral conditions persisting through August 2007, while most dynamical models indicate La Niña will develop within the next three months. However, for the past few months the dynamical models have been predicting a stronger and more rapid cooling than has actually occurred.

6. Summary of the forecasts

i) South-West Monsoon Season Rainfall

IMD's Long Range Forecast update for the 2007 South-West Monsoon Season (June-September) is that for the country as a whole the seasonal rainfall is likely to be 93% of the Long Period Average (LPA) with a model error of ± 4%.

ii) July rainfall

Rainfall over the country as a whole in the month of July 2007 is likely to be 95% of its LPA with a model error of ± 9 %.


iii) Rainfall over broad geographical regions

Over the four broad geographical regions of the country, rainfall for the 2007 South-West Monsoon Season is likely to be 90% of its LPA over North-West India, 98% of its LPA over North-East India, 96% of its LPA over Central India and 94% of its LPA over South Peninsula, all with a model error of ± 8 %.

-------------------------------