Figure 6: Seasonal and geographical variations in normalized total column abundances of water vapor. Panel A: Seasonal distribution of all qualified retrievals averaged in bins of 2°x2° Ls and latitude. Panels B-E: Geographical distribution of qualified retrievals for each seasonal interval, averaged in bins of 2°x4° latitude and longitude.
Given the variation of the sampled local times, it is assumed that seasonal variations of water vapor column abundances dominate over any diurnal variability. Although the temporal and spatial coverage is incomplete, known trends of the water cycle appear unambiguously. The spring is overall a dry season (Figure 6 panel B), with abundances rarely higher than 10 pr-μm. The few instances with larger CIA occur late in spring at the edge of the NPC. The early northern summer season (Ls = 90° –135°) is characterized by large latitudinal contrasts with high water content in the polar regions which decrease monotonically southward, this is particularly prominent in Panel C. At around Ls=115°, a Northern maximum of ∼60 pr-μm is visible poleward of 70°N, while at the same time only around 13 pr-μm is measured south of 30°N (in panel A). The northern hemisphere fall season shown in panel D displays a north polar region now devoid of water vapor, most of which having been transported to the mid and low latitudes and across the equator. The south polar maximum occurs around Ls=285° and reaches an average total column abundance of 40 pr-μm, as seen in Figure 6 panel E. A global dust storm occurred in MY 28, which degraded the quality of the measurements for some time, causing the number of qualified retrievals in MY 28 to be low even though many observations were conducted in this period. The drier patch around Ls=300° in the southern hemisphere is constructed almost entirely from observations in MY 28, when the dust storm is thought to aid transport of water vapor from the lower atmosphere to higher altitudes (Fedorova et al., 2018).
4.2 Synergy compared to single domain retrievals and MCD
Numerous studies of the climatology of water vapor have been made using the PFS and SPICAM instruments individually. As this is the first time observations from both are used in synergy, a direct comparison has been made between them and the MCD. In Figure 7, synergy retrievals and MCD prior values satisfying the adjusted criteria as described in Section 4.1 are plotted, along with single spectral domain retrievals for SPICAM and PFS. No criteria have been imposed on the single domain retrievals other than unphysically high abundances have been filtered out. CIAs are averaged across the 15°-45°N latitude band, the region which contains the longest continuous coverage, and in intervals of 5° Ls.
The selected time period covers the early northern summer, the polar cap sublimation season and continues into late summer of MY 27. In general, the MCD predicts a much higher water vapor abundance than what is obtained with either of the retrieval approaches (except during Ls=120°-140°). The MCD agrees well with the observations only at the very beginning and end of the time period shown here, which corresponds to before the onset of the sublimation season, and after the water vapor has been transported beyond the area of focus. This might suggest that the transportation mechanisms dominant in the summer mid-latitudes are currently not fully understood. Another factor which could impact this discrepancy is the large MCD sublimation peak, which might then propagate southward. The difference in the CIA, as well as the vertical partitioning, predicted by the MCD and the values retrieved by the synergy is further elaborated upon in section 4.3.
The synergy and the single spectral domain retrievals with PFS/TIR are overall in good agreement, with the synergy on occasion yielding slightly lower values. SPICAM/NIR also agrees well with the synergy and PFS, albeit with slightly larger abundances. The general seasonal behavior displayed by the three retrieval approaches is similar; an increasing trend in the early summer, peaking at around Ls=135°, when water vapor from the North Polar Cap (NPC)  has sublimed and been transported to mid-latitudes. The MCD predicts a much more rapid increase of the sublimed water, with CIA values a factor of 2.5 higher than the synergy at Ls=100°. The decreasing CIA found after Ls=140° by both the synergy and the single spectral domain retrievals as well as the MCD is expected, as the water is successively transported across the equator. The “double-hump” shape of the MCD abundances (also evident in Figure 10) are not clearly distinguishable from either the retrieval techniques.