Abstract
Introduction
Data & methods
Results
Conclusion
Acknowledgements
References
 

Main page

Forum

Buoy tracks and AVHRR images were utilised in association in this section to characterise the eddies present in the South-western Atlantic Ocean during the period of March 1993 to July 1994. The two data sets have indicated that there is no relationship between the temperature of the eddies (measured along their perimeter or across their major and minor axes) and their sizes.

The eddy sizes, however, were smaller in the SBCS region than in the BMC region. Distinct properties and generation mechanisms characterise the eddy field in these two regions. Shelfbreak eddies and eddies caused by current reversals occur at the SBCS and at the shelf break off Brazil. To our knowledge, they have not been previously described before. That can be explained mainly because (1) the BC/BCC front, where some of the eddies occur, is still not fully studied and (2) because they are difficult to be spotted in satellite images. However, their presence can be revealed in surface drifters and in satellite imagery if correct enhancing techniques are applied. As mentioned before, depending on the rate and number of structures formed, the shelfbreak eddies of the BC/BCC front can be responsible for a considerable part of the mass and heat exchange between tropical and coastal waters.

In the BMC region, cold core eddies were observed to be formed by breaking off from the crests of high amplitude meanders of the South Atlantic Current. Warm core eddies were present at the BC reversal zone and in the troughs of the SAC in the confluence.

Rotational periods of the small-scale eddies were of about 2 days, while the mesoscale ones (class 2 eddies) presented a mean rotational period of about 17 days. The mean diameter for the mesoscale eddies found in the buoy trajectories is in agreement with those previously described for the BMC region (Legeckis and Gordon, 1982; Garzoli, 1993), but the diameters found for the AVHRR eddies were smaller than previous descriptions. Reasons for that can lie in the fact that the generally bigger warm core eddy formed by the detachment of the BC extreme position, to the south of 42 ^oS, is not detected in our AVHRR images which do not cover that area.

Rossby numbers (Ro) computed for the eddies in the study area indicate very non-linear to quasi-geostrophic or geostrophic structures, these last associated to the SAC in the BMC. The absolute values of Ro for the BMC are in agreement with those presented by Chassignet et al. (1990).

Empirical relationships were found in an attempt to relate the eddy sizes (perimeter and diameter) with the eddy rotational periods (T_R) and tangential velocities (V_T). A better linear adjustment is found between the eddy sizes and T_R (r = 0.65 to 0.79) than between the eddy sizes and V_T (r = 0.39 to 0.4). However, the validation of the models with actual measurements has demonstrated that the percentile errors are much smaller when estimating V_T than when estimating T_R. The relationships can be useful when no in situ data are available and satellite images are obtained in a regular basis. In this case, individual eddy sizes (diameters, perimeters) can be measured in the satellite images and their T_R or V_T inferred from the empirical relationships. The inferred variables could be used as input parameters for ocean models or useful for the Merchant navy, for example, as rotating eddies are know to slow down cruise ships and, therefore, have an economic impact for the navy.
 
 

rbds@soc.soton.ac.uk (c) April 2001, all rights reserved.