A numerical study is made of a time-dependent turbulent Ekman bottom boundary layer. Parameters for the model were chosen to simulate conditions near the bottom of the Florida Current in the Straits of Florida. The model used is that of Lykosov and Gutman. It allows the coefficient of turbulent viscosity v to vary with time t and height z and permits the effects of an imposed stable stratification and sloping bottom to be included. The variation of v with t and z is not preset but is determined in the course of solving the problem subject to the turbulent energy equation, the similarity arguments of Komolgoroff, and a mixing length hypothesis of Zilitinkevich and Laykhtman. The results of this preliminary study are compared to the author's observations. The agreement is good for the friction velocity values as well as for the mean total Ekman veering. However, most of the computed Ekman veering occurred above the logarithmic layer while most of the measured veering occurred within the logarithmic... Abstract A numerical study is made of a time-dependent turbulent Ekman bottom boundary layer. Parameters for the model were chosen to simulate conditions near the bottom of the Florida Current in the Straits of Florida. The model used is that of Lykosov and Gutman. It allows the coefficient of turbulent viscosity v to vary with time t and height z and permits the effects of an imposed stable stratification and sloping bottom to be included. The variation of v with t and z is not preset but is determined in the course of solving the problem subject to the turbulent energy equation, the similarity arguments of Komolgoroff, and a mixing length hypothesis of Zilitinkevich and Laykhtman. The results of this preliminary study are compared to the author's observations. The agreement is good for the friction velocity values as well as for the mean total Ekman veering. However, most of the computed Ekman veering occurred above the logarithmic layer while most of the measured veering occurred within the logarithmic...