Comparisons of modeled and measured ionospheric characteristics are made for data recorded during the January 1993 10-day world day campaign. The peak of the F region hmF2 measured by Digisondes at Ramey (located 50 km west of Arecibo) and at Puerto Madyrn, which is located close to the conjugate point, are used to drive the field line interhemispheric plasma (FLIP) model. In the summer hemisphere for times when the geomagnetic conditions are quiet, the modeled and measured NmF2 are in good agreement, displaying median differences of approximately 13%. In the winter hemisphere, good agreements in NmF2 are observed during midday and evening hours. However, FLIP underestimates NmF2 values at night and in the postsunrise sector. Comparison of the Arecibo ISR and FLIP parallel velocities also show good agreement during daytime, while large differences are observed at night. The discrepancies in the Arecibo/Ramey measurements and FLIP modeled results, at night, is related to the effect of a 4-hour periodic variation in hmF2 and NmF2. Further analysis identifies, for the first time, this 4-hour variation and discusses its characteristics as being consistent with the effect of a 4-hour tide. The signature of the 4-hour periodic temporal variation in the height of the ionosphere is only clearly observable at F region heights during the winter months and at low solar activity. The 4-hour periodic temporal variation is only observed at the tropical latitude stations Arecibo/Ramey and is not observed at the midlatitude station Puerto Madryn. A 4-hour tidal component is not present in the MSIS86 model which FLIP uses for its neutral concentration. Therefore, when tidal forcing is moving the ions and neutrals to lower altitudes, FLIP artificially calculates large recombination rates. The loss in ionization increases and to maintain the nighttime ionosphere the downward flow of ionization also increases.¿ 1997 American Geophysical Union