182
W.B. Rossow
revised theory of absorption line shapes that appears to explain the continuum absorption and
its temperature dependence over both the infrared and microwave portions of the spectrum.
Work remains to confirm the accuracy of this theory and to determine whether it can explain
the continuum absorption at all wavelengths.
Quantifying the magnitude of vertical water vapor transport by different types of atmospheric
motions is important for three reasons: (1) despite the much lower absolute abundance of water
vapor in the upper troposphere, its effect on the surface radiation balance (the greenhouse
effect) is almost as large as that of water vapor near the surface (Hansen et al., 1984; Arking,
1995), (2) the upper tropospheric humidity controls the occurrence of upper level cloudiness,
particularly cirrus, that can have significant effects on the climate (e.g. Randall et al., 1984;
Slingo and Slingo, 1988), and (3) transport of water out of the tropic oceans and into land areas
and higher latitudes depends on the amount of water vapor injected into the upper tropical
troposphere from the ocean surface (e.g. Del Genio et al., 1991). Uncertainties surrounding
the first issue led to some controversy about the sign of water vapor feedback on climate
change (Lindzen, 1980; Betts, 1990; Del Genio et al., 1994 and references therein). Although
current observations suggest that seasonal warming of the upper troposphere in midlatitudes
is accompanied by an increase in water vapor abundance as expected (e.g. Rind et al., 1991);
nevertheless, the complexity of the water vapor budget even in models (Figure 8.6) suggests
the need for much better observations of the space and time variations of water vapor in the
upper troposphere to diagnose the processes affecting its distribution (Del Genio et al., 1994).
In particular, observations are needed of water vapor vertical profiles extending into the lower
stratosphere in both clear and cloudy conditions.
Analysis of RAOBS measurements of water vapor have already provided a lot of detail about the
global and even vertical distribution of water vapor and its variations and horizontal transports
(Peixoto and Oort, 1992). To motivate the need for better observations, estimates of the
variations of water vapor abundance from RAOBS can be used to show the effects of current
measurement uncertainties. The largest variations of column water vapor abundance are with
geographic location and season: Figure 8.7 shows the estimated variations. From the warmer
equator to the colder poles, the annual mean water vapor abundance decreases by more than a
factor of ten; however, in many locations the amplitude of the seasonal variations is as large as
the annual mean value. Estimated uncertainties in even the annual mean values (e.g. Elliott and
Gaffen, 1991) are about 10-20% of the mean values, which is smaller than the average seasonal
variations. The relationship of seasonal water vapor and temperature variations demonstrates
that the water vapor abundance is significantly altered by the large scale circulation (Stephens,
1990; Gaffen et al., 1992), particularly in the tropics, and cannot be treated as a simple function
of temperature, as is still done in many analyses.
Although the seasonal variations of water vapor are probably reliably measured, estimated
variations of water vapor from day to day and with longitude (Figure 8.8) are estimated to
be only about 20-25% and 10% of the mean value, respectively (Peixoto and Oort, 1992),
both comparable to the possible measurement uncertainties. This, in turn, makes it difficult
to determine the magnitude of water vapor transports by wave motions in the atmosphere.
Since the mean winds and water vapor abundances are well determined, the average flux of
water vapor from east to west and from south to north can be determined (Figures 8.9a and
8.10a). The significant features of the east-to-west water vapor transports by the mean winds
highlight the general westerly flows at midlatitudes and easterly flows at low latitudes and show
the largest fluxes occur over the central ocean basins from the warmer western ocean currents
to the colder eastern currents (Figure 8.9a). The blocking action of the continents, especially
by high mountain ranges is readily apparent. A significant intraseasonal variation of westward
water vapor flux at low latitudes is associated with the seasonal monsoons of Asia (Chen et
W.B. Rossow
revised theory of absorption line shapes that appears to explain the continuum absorption and
its temperature dependence over both the infrared and microwave portions of the spectrum.
Work remains to confirm the accuracy of this theory and to determine whether it can explain
the continuum absorption at all wavelengths.
Quantifying the magnitude of vertical water vapor transport by different types of atmospheric
motions is important for three reasons: (1) despite the much lower absolute abundance of water
vapor in the upper troposphere, its effect on the surface radiation balance (the greenhouse
effect) is almost as large as that of water vapor near the surface (Hansen et al., 1984; Arking,
1995), (2) the upper tropospheric humidity controls the occurrence of upper level cloudiness,
particularly cirrus, that can have significant effects on the climate (e.g. Randall et al., 1984;
Slingo and Slingo, 1988), and (3) transport of water out of the tropic oceans and into land areas
and higher latitudes depends on the amount of water vapor injected into the upper tropical
troposphere from the ocean surface (e.g. Del Genio et al., 1991). Uncertainties surrounding
the first issue led to some controversy about the sign of water vapor feedback on climate
change (Lindzen, 1980; Betts, 1990; Del Genio et al., 1994 and references therein). Although
current observations suggest that seasonal warming of the upper troposphere in midlatitudes
is accompanied by an increase in water vapor abundance as expected (e.g. Rind et al., 1991);
nevertheless, the complexity of the water vapor budget even in models (Figure 8.6) suggests
the need for much better observations of the space and time variations of water vapor in the
upper troposphere to diagnose the processes affecting its distribution (Del Genio et al., 1994).
In particular, observations are needed of water vapor vertical profiles extending into the lower
stratosphere in both clear and cloudy conditions.
Analysis of RAOBS measurements of water vapor have already provided a lot of detail about the
global and even vertical distribution of water vapor and its variations and horizontal transports
(Peixoto and Oort, 1992). To motivate the need for better observations, estimates of the
variations of water vapor abundance from RAOBS can be used to show the effects of current
measurement uncertainties. The largest variations of column water vapor abundance are with
geographic location and season: Figure 8.7 shows the estimated variations. From the warmer
equator to the colder poles, the annual mean water vapor abundance decreases by more than a
factor of ten; however, in many locations the amplitude of the seasonal variations is as large as
the annual mean value. Estimated uncertainties in even the annual mean values (e.g. Elliott and
Gaffen, 1991) are about 10-20% of the mean values, which is smaller than the average seasonal
variations. The relationship of seasonal water vapor and temperature variations demonstrates
that the water vapor abundance is significantly altered by the large scale circulation (Stephens,
1990; Gaffen et al., 1992), particularly in the tropics, and cannot be treated as a simple function
of temperature, as is still done in many analyses.
Although the seasonal variations of water vapor are probably reliably measured, estimated
variations of water vapor from day to day and with longitude (Figure 8.8) are estimated to
be only about 20-25% and 10% of the mean value, respectively (Peixoto and Oort, 1992),
both comparable to the possible measurement uncertainties. This, in turn, makes it difficult
to determine the magnitude of water vapor transports by wave motions in the atmosphere.
Since the mean winds and water vapor abundances are well determined, the average flux of
water vapor from east to west and from south to north can be determined (Figures 8.9a and
8.10a). The significant features of the east-to-west water vapor transports by the mean winds
highlight the general westerly flows at midlatitudes and easterly flows at low latitudes and show
the largest fluxes occur over the central ocean basins from the warmer western ocean currents
to the colder eastern currents (Figure 8.9a). The blocking action of the continents, especially
by high mountain ranges is readily apparent. A significant intraseasonal variation of westward
water vapor flux at low latitudes is associated with the seasonal monsoons of Asia (Chen et