The weather leading to the recent floods and landslides in Sri Lanka 

Lareef Zubair
 Sri Lanka Meteorology, Oceanography and Hydrology Network and  
International Research Institute for Climate Prediction, Earth Institute at Columbia University, New York, USA.

lareef@iri.columbia.edu

From the 11th to the 19th of May 2003, a tropical storm made its way across the Bay of Bengal causing perhaps 300 deaths, the displacement of 200,000 persons, and heavy damage to the infrastructure, economy and livelihoods of the South-Western Sri Lanka. This  storm first formed 700 km to the West of Sri Lanka on May 11 and  then made its way North by about 500 km. It stalled there for a week.  On the 14th of May, the wind speeds were high enough for it to be classified as a cyclone for a day, there after the wind speeds declined. 900 km away frmom the eye of the cyclone, there was heavy rainfall over the South-Western corner of Sri Lanka. This cyclonic storm eventually reached Burma on the 23rd of May and dissipated thereafter.

 May is the month when the heavy Yala season rains start and the South-West of Country is expected to become wet. Indeed, there have been heavy rainfall of 600 mm in Sri Lanka during the May from 1920 to 1940 but not thereafter.  Of course, there have historically been no record of a cyclone bearing on Sri Lanka during May. It was also rare that a cyclonic storm whose centre was 700-1500 km away and which was relatively weak cyclone could cause such damage.

Overall, the disaster resulted from a rare combination of three factors: 
    1. Intense South-Easterly Winds in the Bay of Bengal that stalled the cyclone in the middle of the Bay of Bengal and also contributed to the high winds in the South-West of Sri Lanka
    2. The cyclone being stalled in approximately the right location that woudl induce high wind speeds over Sri Lanka
    3. This event taking place in the middle week of May when the Tropical Convergence Zone band of clouds hovers over Sri Lanka.

All of this resulted in high winds near the South-Western coast. The component of these winds that is directed towards the North-East gets obstructed by the topographic features such as the Adam's Peak and Gongala mountains. As is usually the case, with mountain induced rainfall, the maximum rain fell to the windward side of the mountain.  This obstructions lead  to high rainfall in the South-Western corner of Sri Lanka every year during this period. What was different that fateful week in May was that the winds were much more intense than is usual. It also took place during a point in which there are cloud bands located over Sri Lanka ready to precipitate.



The map below shows the rainfall for the week leading into the worst damage. Thus already the Western Plains region was wet. The rains in the following weeks were twice as intense and located pretty much in the same area. The storm abated by the 21st as it made landfall in Burma.  

Rainfall Distribution in Sri Lanka from May 6-13, 2003
 


 
In a historical sense, the monthly rainfall for May is high (600 mm) but is not the highest on record in the South-Western region as shown in the graph below.  But the distribution of sustained high rainfall  in just two weeks led to the heavy flooding and landslips.

Historical Rainfall in South-West for May from 1869 to 1998
Color Scale


There were floods and cyclones leading to much of the damage.

Disaster Information from Relief Web


Cyclones usually are steered towards Sri Lanka only during the North-East Monsoon period and we only expect Cyclones to make landfall during this period. So the forecasters were perhaps caught off-guard. Indeed, this cyclone did not make landfall in Sri Lanka.

Seasonal Distribution of Cylonic Storms Making Landfall in Sri Lanka


This was really a cyclone that tracked its way across the Bay of Bengal at least 750 miles away from Sri Lanka.  

Actual Track of the Cyclone

There was only a single day when the winds reached high enough speeds (May 14, 2003) so that it could the storm could be called a cyclone.  


The map below shows the intense South-Easterly Winds that stalled the Cyclone for a week and also the cyclonic pattern in the Bay of Bengal. .
     
Winds in the Bay of Bengal on the 14th of May, 2003


The wind vectors in the Bay of Bengal were in a South-Easterly direction and as a result the eye of the cyclone howered around th e same place for a week. The particular location of Sri Lanka led to an intensification of wind near Sri Lanka.


High Resolution Winds on the 11th of May

    
These high winds during the particular weeks in May in which the Tropical Convergence Zone cloud bands hover over Sri Lanka - this is likely to be one reason why there was such high rainfall in Sri Lanka.

The component of these winds that is directed towards the North-East gets obstructed by the topography of Sri Lanka (see below). This obstructions lead to the orographic rainfall mechanism that year-in year out causes high rainfall in the South-Western corner of Sri Lanka. What was different that week in May was that the winds were extremely high.

Topography of Sri Lanka  

The regional pattern of rainfall as recorded by the TRMM Satellite Observations were:

The Weekly Regional Rainfall Patterns over the Bay of Bengal
 
 
Note that the heaviest rainfall took place in Sri Lanka from the 17th until the 20th of May.


Mail Queries to Lareef Zubair at lareef@iri.columbia.edu