454
P. Schliissel
Figure 16.10: Variation of near-surface specific air humidity with total columnar water vapour
content and with the water-vapour column of the lower 500 m in the time period July 1987 to
June 1988 over the global ocean.
technique of Liu and Niiler (1984) is used with the water-vapour columns w as derived from
SSM/I channels at 22 Ghz and 37 GHz, vertical polarization, according to Schliissel and Emery
(1990). Also, a technique is incorporated in determining q by fixing the relative air humidity
to 80% and assuming the air temperature to be equal to the sea surface temperature. This
method only requires the knowledge about the sea surface temperature for the retrieval of q.
It is seen that the direct method according to (16.31) gives the lowest systematic and standard
deviations with respect to the ship measurements of q. At high latitudes, the 80% - method
gives similar results as the other techniques. In other latitudes great systematic errors must be
expected when following this simple approach. However, it should be kept in mind that one
of the main error sources in such intercomparisons is the temporal and spatial variability of
water vapour between satellite measurements and ship observations that have been matched
for compiling the statistics. Here, the time and space windows used cover differences of one
hour and 25 km, respectively.
P. Schliissel
Figure 16.10: Variation of near-surface specific air humidity with total columnar water vapour
content and with the water-vapour column of the lower 500 m in the time period July 1987 to
June 1988 over the global ocean.
technique of Liu and Niiler (1984) is used with the water-vapour columns w as derived from
SSM/I channels at 22 Ghz and 37 GHz, vertical polarization, according to Schliissel and Emery
(1990). Also, a technique is incorporated in determining q by fixing the relative air humidity
to 80% and assuming the air temperature to be equal to the sea surface temperature. This
method only requires the knowledge about the sea surface temperature for the retrieval of q.
It is seen that the direct method according to (16.31) gives the lowest systematic and standard
deviations with respect to the ship measurements of q. At high latitudes, the 80% - method
gives similar results as the other techniques. In other latitudes great systematic errors must be
expected when following this simple approach. However, it should be kept in mind that one
of the main error sources in such intercomparisons is the temporal and spatial variability of
water vapour between satellite measurements and ship observations that have been matched
for compiling the statistics. Here, the time and space windows used cover differences of one
hour and 25 km, respectively.