Peixoto, J. P., and A. H. Oort, 1996: The climatology of relative humidity in the atmosphere. Journal of Climate, 9(12), 3443-3463.

Abstract: The present paper deals with the analysis of the time-average relative humidity fields in the atmosphere. Twice-daily estimates of relative humidity are used.

After some theoretical considerations on the relations between relative humidity, other moisture parameters, and temperature, a critical analysis of the various sources of data is made considering their possible limitations. Various methods of computing relative humidity are formulated and discussed.

The global distribution of relative humidity at various levels shows that it is not zonally uniform with centers of various intensities at all latitudes. The global maps show maxima in the equatorial zone and minima in the dry subtropical belts around 30°N and 30°S. The land-sea contrast and variations related to the orographic relief are also apparent. The general pattern of relative humidity is similar at all levels but its magnitude decreases with altitude. The seasonal analyses show a similar pattern as the annual analyses but are slightly shifted toward the summer pole.

The saturation deficit is also evaluated. Cross sections of the saturation deficit show that the maxima are found in the middle to lower troposphere at subtropical latitudes, being most intense in the Northern Hemisphere during the summer season.

The temporal variability of the relative humidity due to transient eddies exhibits a bimodal structure with maxima in the midlatitudes of each hemisphere around 700 mb. The stationary eddy distributions are less pronounced than the transient ones and do not change substantially from one season to another.

To validate our results, several comparisons are made using independent sources of humidity data as well as cloud distributions at various levels. Thus, COADS data are used to obtain independent analyses of the surface relative humidity over the oceans, and satellite observations by SAGE are used at the 300-mb level. The rawinsonde-SAGE differences are on the order of 10% in the Tropics and 20% in the high latitudes, due in part to a clear-sky (dry) bias in the SAGE data. Our results are further compared with those obtained from operational analyses by the ECMWF. The differences do not exceed 5% in the Tropics but tend to be larger in the tropical upper troposphere and at all levels in the extratropics of the Southern Hemisphere, where the radiosonde network is quite sparse. In view of the obvious connections between the moisture distribution in the atmosphere and cloudiness, a cloud climatology is used to check its consistency with the present results. The latitudinal and interseasonal variations of cloudiness and relative humidity are similar, with maxima in the equatorial belt and at high latitudes and minima in the subtropics that shift poleward during summer and equatorward during winter.

Finally, some comments are made on the radiosonde-observing systems in the light of recent satellite studies of humidity. Mainly at the upper levels systematic localized differences are found between electrical hygristor and organic sensors, but the differences almost disappear in the middle and lower troposphere.

In spite of the shortcomings, limitations, and errors of the radiosonde network, the present analyses describe for the first time the large-scale, three-dimensional characteristics of the relative humidity in the global atmosphere.