The ongoing drought in the Hambantota District in the South-East of the Island has led to shortages of drinking water, failure of crops, malnutrition, typhus and hardship to wildlife. There has been an call for drought assistance internationally. In this article, I answer some questions on the climatological background to this drought.

However, a word of caution - I am writing far away from Hambantota, and relying on station and satellite data as reported in various archives. One may cross check the older data, but not the current data. Hence, you should be careful about the inferences based on on the recent data.

1. Is there indeed a drought?

Yes, the available data shows drought conditions. I do not have rainfall data for Hambantota, but consider the figure below showing rainfall at Nuwara Eliya. The bottom panel shows the rainfall as reported by the Department of Meteorology. The top panel shows both the cumulative normal rainfall for 365 days starting on September1 (dotted line) and the actual cumulative rainfall during the last 365 days (solid line). The extent of the brown shading shows the cumulative rainfall deficit. The dates on which the data were missing are shown as blanks.

From September 2000 to this August 2001, the observed rainfall at Nuwara Eliya (1350 mm) has been below the normal rainfall (2200 mm) by 40%. This is a pattern seen right through Sri Lanka and in particular in Southern extreme of the island.

2. What is the regional distribution of drought?

The rainfall during the three months from May to July  2001 has been estimated by the US NOAA/CPC based on satellite estimates and station data. The graph below shows the rainfall that fell in any region as a percentage of what would be normal for that time and place. Do scroll to the right to see the scale. It shows that the rainfall in the Southern extreme of Sri Lanka had dropped by 25-50% from normal years. This drop compounded the shortfall in rainfall that started in February.

Although, there appears to be dark brown patches in the Indian peninsular, bear in mind that this is the summer monsoon season in India with its heavy rainfall and and even 50% of normal rainfall does not necassarily imply drought.  
3. What is the seasonal distribution of rainfall at Hambantota?

Rainfall records have been maintained since 1869 and these archives are now available in the Department of Meteorology and the British Meteorological Agencies.  The monthly average rainfall is graphed below. The main rainfall season starts in September and goes on till January. One can hope now, that rainfall during the upcoming months will alleviate the situation.

The average rainfall at Hambantota is 1016 mm as compared with an island-wide average of 1850 mm.

4. How bad is this drought in the historical context?
In analyzing rainfall in Sri Lanka, it is best to separate out the period of heavy rainfall from October to December. 

The aggregate January to September rainfall at Hambantota from year to year is shown below as a bar chart. The colours of the bar are based on the magnitude of rainfall. The aggregate rainfall ranges from 150 to 1150 mm with an average of 582 mm.  The total for the present year estimated from various sources and newspaper reports is estimated as 419 mm.

On the face of it, this quantum of rainfall is not too low. However, there are three compounding factors. Nearly a third of the rainfall fell during January 29th and the rainfall since then has been paltry. In addition, the rainfall in the hills of the island that feed the various streams that run out to the Sea in the South too have been affected badly by low rainfall in the hills. In addition, the last several months have been quite hot in Hambantota.
 

5. What may we expect of the rainfall from October to December?

The main rainy season in Hambantota starts  in late September and ends in early January. During these months, Sri Lanka garners  rainfall form the "winter monsoon", cyclonic storms form the Bay of Bengal and convectional rains from solar heating of the oceans at the latitude of Sri Lanka. The average rainfall during October to December for Hambantota (434 mm) is quite high and equal to that during the seven months from February to August (437 mm). Since 1869, the aggregate October to December rainfall has ranged from 80 mm in 1964 to 1031 mm in 1912. The lowest was in 1964, however the high rainfall in the preceding January to September period compensated for the low rainfall in the subsequent three months.During 2000, the rainfall during October and November were normal. The rainfall  data for December was not available to me.

6. What are the prospects for rainfall in the immediate future?

Experimental forecasts for 5-10 days are made by Center for Ocean, Land and Atmosphere, Maryland, USA. These forecasts are experimental and I am not aware as to its skill. The prediction for the next 5 days is for no rainfall. However, the rainfall thereafter is predicted to be about 60 mm all over the island.

6. What are the prospects for rainfall in the next months?

Climate predictions can be made based on statistical relations between rainfall and its precursors such as snow cover in the Himalayas or the state of the Oceans. Indeed, the rainfall at Hambantota appears to be affected by El Nino type conditions in the Pacific Ocean and warm Indian Ocean conditions. The Indian Ocean has been anomalously warm this year and may well have contributed to the low rainfall in Hambantota from January to September. However, these same warm oceanic conditions are also related to increased October to December rainfall. Such historical analogies suggest that we should have normal rainfall in the upcoming months.

However, predictions based on three global climate models at the International Research Institute for Climate Predictions (IRI) point towards lower rainfall and higher temperatures from September to February (see figures below). These predictions are based on global climate model simulations with boundary conditions for oceanic and land surface conditions. Such simulations are a new development and are very much under development. In particular, events such as cyclones are not accounted for in these forecasts and can throw off predictions. Remember, that the cyclones reach Sri Lanka most often in November and December.

These contradictory prognostics are confusing but is the current state of the art. The quality of predictions should improve with more  research.

You may email the authour from
http://www.columbia.edu/~lz144/slmon/
using a web-form.