We present observations of the diurnal formation of horizontal temperature gradients in the surface waters of a sidearm of a small water supply reservoir at a time during summer when radiative heating and vertical stratification of the reservoir's waters were both quite strong. Our measurements show that because the closed end of the sidearm was relatively shallow, daytime heating and nighttime cooling created larger temperature changes there than in the body of the lake, resulting in large horizontal temperature gradients that drove strongly sheared, horizontal exchanges. Because of more vigorous turbulent mixing during cooling, cooling-driven flows were slower and of greater vertical extent than were heating-driven flows. The overall flow exhibited inertia in that it was not in phase with the daily heating cycle. Averaged over the daily cycle, there appeared to be a net residual flow, with surface waters flowing out and metalimnetic waters flowing in. This thermally driven flow, the "thermal siphon," greatly enhanced the rate of horizontal exchange between the sidearm and the body of the reservoir such that the time required to replace water in the sidearm when the siphon is operating is substantially less than estimates with conventional formulae based on horizontal turbulent diffusion.In Mediterranean climates, a diurnal cycle of stratification is observed in lakes from the beginning of spring until the end of fall (Imberger and Patterson 1989). During the day, a layer several degrees warmer than the wakers below and a few meters deep forms through the combined actions of surface heating and wind mixing; at night, this layer deepens considerably and cools convectively in association with surface heat losses . Weekly variations in surface fluxes associated with the movements of synoptic weather systems introduce corresponding variations in the depth and heat content of the surface mixed layer, such that the daily repetition of this cycle results in
AcknowledgmentsWe thank Cliff Heam for collecting the drogue data, personnel from the Australian Survey Office for assisting in the drogue position fixing, Greg Ivey and John Patterson for reviewing an early draft ofthe paper, and Dave Pullin, Richard Chapman, Carolyn Oldham, and Terry Smith for assisting in the operation of the equipment on the Djinnang II.