The Indian Meteorological Department (IMD),Pune,has a tough reputation to live up to this year with its annual forecast for the 2009 Southwest Monsoon to be announced next month. In its monsoon report last month,the Ministry of Earth Sciences has claimed that its forecast for 2008-09 was almost bang on target.
According to the long range forecast for the 2008 Southwest monsoon issued on April 16,the seasonal rainfall for the country as a whole was expected to be 99 per cent of the long period average (LPA) with a model error of ± 5 per cent. In the updated forecast issued on June 30,the forecast was revised to100 per cent of the LPA with a model error of ± 4 per cent. The actual area-weighted rainfall for the country as a whole was 98 per cent of the LPA and it made the predictions fairly accurate.
This also bodes well for the IMD model put in place in 2007 for monsoon predictions. At least it shows us that we need not change anything,though its too early to rejoice over the accuracy. The first year there were some errors that we rectified,last year was very close,but we need to see for another two-three years more before we can say anything, said Dr DS Pai,director of Long Range Forecasting,IMD Pune.
The IMD is now working on compiling data from all over the country to prepare the monsoon forecast for 2009 that will be issued in April. BN Goswami,director of India Institute of Tropical Meteorology,also feels the model needs to go through a few more cycles before its efficiency can be ascertained,despite the accuracy achieved last year. The institute had developed the analogues technique for real time prediction of active-break spells during the monsoon 2008 season and the IMD had used the technique for the first time in 2008.
We have a statistical and dynamic model to make seasonal predictions that are sent to the IMD. The dynamic model was used more this time, he said.
The monsoon report also outlined some of the other salient features,including that while the operational long range forecasts issued by the IMD were mostly accurate,the forecast for seasonal rainfall over northwest India and that for July rainfall in the country as a whole were not so.
While the forecast highly overestimated the rainfall during July for the country as a whole,it underestimated the seasonal rainfall over northwest India.
2008 monsoon highlights
* Onset over Kerala on May 31 one day before normal. Covered entire country 5 days earlier than normal date of July 15
* Withdrawal of monsoon from western Rajasthan was delayed. Began September 29 instead of September 1
* Low pressure systems subdued in comparison with earlier years
* Rainfall marked by intra-seasonal variation. In excess in June,deficient in July,near-normal during August and September