FIGURE 8.3 The difference between the surface pressure climatologies for July and January, showing the peak-to-peak variation due to the atmospheric annual cycle in mass. Contour interval is 5 mbar. Blue values are positive, red values are negative. The climatology is a multi-year average for the month in question, based on recent reanalysis at NCEP (Kalnay et al., 1996).

surrounding ocean areas, capable of inducing significant local gravity  anomalies. While concomitant atmospheric pressure perturbations over the oceans would be redistributed more or less uniformly by the inverted barometer effect (see Chapter 4), the large land-ocean mass flux involved in this and other seasonal phenomena may be capable of inducing an average pressure change over the surface of the world oceans that could have a significant impact on the global gravitational field.

On short time scales, intraseasonal variability includes the important effects of cold (blocking) outbreaks over mid-latitude continents (cf. Figure 1.2), and the effects of the eastward-propagating surface pressure waves associated with the tropical Madden-Julian Oscillation. At interannual time scales the El Niño-Southern Oscillation (ENSO) phenomenon dominates global-scale atmospheric variability, with concomitant meridional redistribution of atmospheric wind and mass fields (e.g., Dickey et al., 1992; Mo et al., 1997).

The detection of atmospheric gravitational effects is not likely to be an important goal of time-dependent gravity studies because of the continued generation of high-quality pressure fields by meteorological forecast centers. Although it is possible that satellite gravity constraints on atmospheric pressure over data sparse regions (e.g., parts of the Southern Hemisphere such as Antarctica) could provide results that are more accurate than those available from meteorological analyses and that such results could be used as a proxy data type in reanalysis efforts (e.g., Kalnay et al., 1996), extending barometric networks on the surface would most likely provide the best return on available resources (see Chapter 7).

The better recovery of certain non-atmospheric signals could also greatly enhance atmospheric modeling capabilities. Particularly promising are the hydro-