Seasonally averaged quiet-day F region ionospheric E x B drift observations from the Millstone Hill, St. Santin, Arecibo, and Jicamarca incoherent scatter radars are used to produce a model of the middleand low-latitude electric field for solar m'mimum conditions. A function similar to an electrostatic potential is fitted to the data to provide model values continuous in latitude, longitude, time of day, and day of the year. This model is intended to serve as a reference standard for applications requiring global knowledge of the mean electric field or requiring information at some location removed from the observing radars. 1974; Kohl, 1976], and by influencing the magnitude of ion iations of the electric field, which are not incorporated into drag through redistribution of ionization [e.g., Anderson and Richmond's [1976] model. The present model thus supersedes Roble, 1974]. Knowledge of the global ionospheric electric that earlier one, field is also useful in providing an upper boundary condition on calculations of middle atmosphere electric fields such as those performed by Roble and Hays [1979]. Substantial collections of ionospheric electric field data exist for the incoherent scatter radar observatories at Millstone Hill, St. Santin, Arecibo, and Jicamarca. In spite of the large day-to-day variability of the electric fields, earlier analyses were able to determine clear average daily variations [e.g., Woodman, 1970, 1972; Evans, 1972; Carpenter and Kirchhojf,, 1974; Kirchhoff and Carpenter, 1976; Richmond, 1976; Blanc et al., 1977]. More recent analyses have quantified the seasonal DATA BASE Data are used from the four incoherent scatter radar stations listed in Table 1. These radars measure the ion velocity, whose component v perpendicular to the geomagnetic field B is related to the electric field E by v = E • B/B • (•) At each station, F region ionization drifts were typically measured on a few days each month within the interval indicated. Depending on the mode of operation, projections of v onto one or more station-dependent axes can be determined at a given time. The measured components of v are averaged in height to improve accuracy, a process justified by the expectation that E and v vary only slightly with altitude within the F region, owing to the near-equipotentiality of magnetic field lines and to the large-scale nature of the global electric field. The mean altitude of the measurements is roughly 300 kin. All stations except St. Santin determine vector velocities by combining line-of-sight velocities measured at different directions from the transmitter/receiver, corresponding to different volumes of the ionosphere. The assumption is then made that the ionospheric drift vectors are the same for all volumes of measurement from a particular station. For further information about the measurement techniques, see Paper number 80A0414.