Introduction

Atmospheric pressure measurements from the Viking Landers have been singularly valuable in understanding the various processes taking place in the Martian atmosphere. Although this is partially due to the quantity and quality of the measurements, it is mainly due to their relevance to scales from local to global. Previous studies have treated synoptic processes, atmospheric tides, global oscillations, the determination of optical depth, and the annual variation due to polar cap condensation and sublimation [Barnes, 1981; Hess et al., 1977, 1979, 1980; James and North, 1982; Leovy and Zurek, 1979; Leovy, 1981; Leovy et al., 1985; Paige and Wood, 1992; Ryan et al., 1978; Ryan and Henry, 1979; Tillman, 1977, 1985, 1988; Tillman et al., 1979; Wood and Paige, 1992; Zurek, 1981, 1988; Zurek et al., 1992]. The desire for an accurate mean pressure, and a spectral model of the annual variation, led us to develop this description of the annual cycle in terms of the mean, the fundamental and the first four harmonics of the annual $\rm CO_2$ condensation-sublimation cycle for the Viking Lander sites. Here we report the results for a two year period selected to exclude `` great'' dust storms, and establish that these two years are almost identical at Lander 1 and very similar to the Lander 2 data coincident with the first Lander 1 year.

During the first year of the Viking mission, two major dust storms were observed (Figure 1), the 1977 A and the 1977 B [Tillman, 1988]. Whether or not the second year had a global or major dust storm of intensity similar to the weaker 1977 A storm has been the subject of debate. For example Leovy [1981] concluded that the second year has only one major storm, suggesting that `` Lander 2 tides follow an almost identical pattern during the period from $L_S = 180^\circ$ to $L_S = 260^\circ$ $\ldots$ with the same semidiurnal amplitude but with a somewhat smaller diurnal amplitude than during year 1.'' ( L<sub>S</sub> represents the aerocentric longitude of the sun and is a seasonal index, $L_S\; 90 =$ summer, $L_S\; 270=$winter.) However, this is qualified by `` Lander 1 data are too sparse to define this event well, but the semidiurnal amplitude observed at $L_S = 230^\circ$ suggests that it was distinctly less intense than 1977a.'' In other analyses [Leovy et al., 1985; Tillman, 1985] it is concluded that great dust storms occurred only in years 1 and 4.