Most of the available hydrographic data (1939--1988) from the Gulf of California are used to describe the seasonal cycle of heat storage in the upper 400 m of water. The seasonal cycle of surface heat flux is obtained by using the sea surface temperature from the hydrographic data bank and meteorological data from coastal stations along the gulf. Monthly values of surface heat flux and heat storage are fitted to a mean value plus an annual harmonic. The longitudinal (i.e., along-gulf) heat flux is then obtained by integrating the difference between the rate of change of heat storage and the surface heat flux. It is found that the surface heat flux has a positive (into the sea) annual mean all along the gulf; the average annual net surface heat flux for the whole gulf being ~118 W m-2. The southern part of the gulf gains more heat (~130 W m-2) than the northern part (~100 W m-2), owing to the larger loss by evaporation in the latter (~0.6 m/yr versus ~1 m/yr) caused by lower humidity and stronger winds.

These values are in agreement with previous estimates for the northern gulf. The amplitude of the seasonal signal shows a maximum of ~230 W m-2 in the central part. Winter heat losses occur only from the middle of the gulf, the Guaymas Basin, to its head. There is a strong annual signal superimposed on the average longitudinal heat flux, whose amplitude is large than the annual harmonic of the surface heat flux. Input of heat from the Pacific Ocean occurs from mid-March to mid-July, with a maximum at the mouth in May (21¿1012 W). Attempts to explain this flux by diffusive processes were unsuccessful. The proposition that advection is the main longitudinal heat carrier is congruent with the generally accepted circulation patterns. ¿American Geophysical Union 1994