Behaviour of oceanic mesoscale eddies over the Bay of Bengal in contrasting monsoon years

Subhra Prakash Dey (IIT Kharagpur, India, India)


Mihir K. Dash (IIT Kharagpur, India); P. C. Pandey (IIT Bhubaneswar, India)

Event: 2014 Ocean Surface Topography Science Team Meeting

Session: Science Results from Satellite Altimetry: Finer scale ocean processes (mesoscale and coastal)

Presentation type: Type Oral

Oceanic mesoscale eddies are mainly formed due to the instabilities in ocean currents. If there is any disturbance in ocean circulation that must be reflected in these mesoscale eddies. It is well known that there are two types of monsoonal circulation (north-east and south-west monsoon) over the Indian subcontinent and that affect the ocean circulation hence the mesoscale eddies. In this study we mainly focused on the various properties of these eddies and their variation over Bay of Bengal (BoB) in contrasting monsoon years (drought years 2009 and 2002 and normal years 2005 and 2003) and the possible mechanism for different behaviour in contrasting monsoon years.
For this study we used AVISO daily geostrophic velocity product (spatial resolution 0.25° × 0.25°) and weekly AVISO sea level anomaly (SLA) data (spatial resolution 0.25° × 0.25°). Both of these are merged products of any four satellites of Jason-2, Altika, Cryosat-2, Jason-1, Envisat,Topex/Poseidon, ERS at the given time. Weekly SLA data is interpolated to daily using spline interpolation technique to study the intraseasonal oscillations in SLA.
We use Okubo-Weiss method (Okubo; 1970) method we calculated the vorticity dominated regions to delineate the cyclonic and anti-cyclonic eddies. Then the areas of these eddy regions and total kinetic energy over the regions for both the eddiesare calculated separately (according to the sense of rotation or sign of the vorticity). We divided the whole Bay of Bengal (78°E - 99°E and 5°N - 23.5°N) into three regions along the latitude; the southern most part is 5°N to 11°N, mid bay region is 11°N to 17oN and the northern part is 17°N to 23°N. Daily variation of total area and total kinetic energy for cyclonic and anti-cyclonic eddies is calculated separately in the three domains for pre-monsoon (March - May) and monsoon (June - September) (Figure 1 and 2).The analysis shows that total area and total kinetic energy of both cyclonic and anti-cyclonic eddies in 2005 is larger than those in the 2009 for almost all the study period. However,although cyclonic eddy kinetic energy in 2002 is always less than that of 2005, cyclonic eddy area in 2002 is not always less than that of 2005 in pre-monsoon time. In the monsoon time both kinetic energy and eddy area are almost always large in 2005 than those of 2002. The anti-cyclonic eddy area is larger in 2005 than that of 2002 from first April to last week of June. There is no significant contrast found in the variation of eddy area in 2002 and 2003. In other two domains of the BoB there is no distinct variation found in the eddy properties in contrasting monsoon years. Hence eddy activity over the northern part of the Bay is much more sensitive to monsoon circulation.
We also tried to find out the possible mechanism for these variations in eddy energy and eddy area in contrasting monsoon years. Intraseasonal oscillations (30 - 100 days) in SLA mainly reflectinedthe eddies. In figure 3(a) and 3(b) the standard deviation of the 30 - 100 days filtered SLA is plotted for 2005 and 2009 respectively. These two figures show that at the Sumatra coast and at the eastern rim of BoB standard deviation is much higher in 2005 than that of2009. These oscillations travel along the coast of BoB as coastal Kelvin waves and radiate Rossby wave along the west. These westward propagating Rossby waves mainly represent through eddies activities in this region. Since in 2009 intraseasonal oscillation along the eastern BoB was much less than that of 2005. It signifies that the strength of coastal Kelvin wave was less, hence the radiated Rossby wave was also less. To examine the propagation of Kelvin wave we considered 69, 1°×1° boxes from (0°, 80°E) long equator to the Sumatra coast (equatorial waveguide) and then along the rim of the BoB. Then we regressedintraseasonalSLA in pre-monsoon time at (0°, 90°E) box with that at the other boxes with lead lag 45 days and then plotted the regression coefficient with horizontal axis as box numbers and vertical axis as lead days. For 2005 propagation of the wave is clearly found but for 2009 and 2002 the strength of Kelvin wave along the rim of BoBwas found to be reduced. In 2002 though there were oscillations at the Sumatra coast and the eastern rim of BoB, the disturbance did not propagate at that year. This might be the reason for less eddy activityin the year 2002.

Oral presentation show times:

RoomStart DateEnd Date
Red salon Wed, Oct 29 2014,15:30 Wed, Oct 29 2014,15:45
Subhra Prakash Dey
IIT Kharagpur, India