VOL. 76, NO. 18 JOURNAL OF GEOPHYSICAL RESEARCH JUNE 20, 1971 Monsoonal Response in the Western Indian Ocean WALTER Di3I•a Rosenstiel School of Marine and Atmospheric Science University o• Miami, Miami, Florida 33149 •ARL-HEINZ SZEKIELDA 2 Goddard Space Flight Center, Greenbelt, Maryland 20771 Infrared observationsfrom spacecraft are used to investigate the responseof the Somali Current to the onset of the southwest monsoon. Selected satellite observations from three years were available for this study (Nimbus series1966,1969,and 1970). The time-dependent development of horizontaltemperature gradients at the sea surface serves as an indicatorfor the formation of the baroelinie structure of the Somali Current. A comparison is made with the simultaneous development of the southwest componentof the monsoon wind. The investi- gationreveals that the temperature gradients duringthe early formation stage in all years are directly proportionalto the wind speed.The phase lag between the development of wind and temperaturegradient during the buildup of the boundary current has a mean value of twelvedays.During the decay period in late summer and fall, the lag increases continuously up to forty days. The observations suggest that two phenomena of differentspatial scales play an important role during the formation of the Somali Current;in the early stage (May, June),local wind-induced upwelling seems to be the more important source of baroelinieity; in the later stage of the buildup(July), large-scale geostrophie effects seem to be dominating. INTRODUCTION Very little is known about the response of western boundarycurrentsto large-scale fluc- tuations of wind systems. Two main reasons account for the almost complete lack of such ob- servations for midlatitude boundary-currentsys- tems such as the Gulf Stream and the Kuroshio. The occurring fluctuations are of smallampli- tudecompared to themean transport, and their periods seem to be in the order of months, or even years. For example, the total transport fluctuations of the Florida Current at nontidal periods do not seem to exceed 10% (3 X 106 mS/sec) of the mean transport [Schmitz and Richardson, 1968]. A more suitable objectfor the observation of x Contribution 1347 from the Rosenstiel School of Marine and Atmospheric Science, University of Miami. ß On leave from the Facult.• des Sciences,Mar- seille, France, as a National Academy of Sciences- National Research Council Postdoctoral Resident Research Associate. Copyright • 1971 by the AmericanGeophysical •lnion. boundary-current response to atmospheric forc- ing is the Somali Current,which flowsalong the coast of East Africa duringthe summer. Each year, in response to the onset of the south- west monsoon, it builds up from essentially zerotransport to approximately 60 X 106 m3/sec. The large change of transport,and its annual repetition, facilitate studies of time-dependent behavior. A precise determination of the time lag be- tween the onset of the southwest monsoon and the development of the Somali Current is of great interestto dynamical oceanography. Un- fortunately, no historical recordsexist from the area that are suitable for determining this pa- rameter. Even the rather extensive efforts of the International Indian Ocean Expedition failed to produce such information [Diiinq, 1970]. Various atlases, based on ship drift re- ports, presentaverage data over large time and space intervals. They indicate that the reversal from weak southwestward flow duringwinter to strong northeastwardflow during summer oc- curs within a time span of less than two months. A more precise determination of the nature of 4181