510 JOURNAL OF APPLIED METEOROLOGY VOLUME7X from 1.12-6.21 at the 0000 GMT observation and1.30-4.25 at the 1200 GMT observation.3. Discussion Surface moisture becomes a poor predictor of liquidequivalent in depth in the warm season, particularly atlow latitudes in the United States in the morning, forthe following reasons: 1. Climatologically, the major influx of moisture isat low levels. 2. A mechanism for distribution in depth is neededfor high correlation. Mos~ of the mechanisms such aslow pressure systems, frontal and prefrontal convergence, stay north of the San Antonio area most of thetime in ~ummer. Insolation as a mechanism for verticaltransport of moisture has a greater effect on the 0000GMT than on the 1200 GMT sounding. 3. Homogeneity of low-level air masses in southeastTexas in summer contributes to a relatively small rangein surface moisture. 4. The lack of surface heating at the 1200 GMT observation time tends to suppress vertical mixing, thereby restricting high moisture values to the surface layer.This, in turn, results in surface dew points unrepresentative of moisture in depth. In more northerly latitudes and in the cooler seasonthe greater frequency of low pressure systems contributes to vertical transport of low-level moisture plusadvection of higher level moisture, both of whichordinarily contribute to higher correlations. In addition, a larger range with equal scatter in surface dewpoints also contributes to higher correlations. On particular days, however, the relationship of surfacemoisture to liquid equivalent in depth may be quitepoor.4. Conclusion Even with a judiciously chosen estimation of meanvertical distribution of moisture, one should expect,for some areas and seasons at least, some large errorsin trying to estimate liquid equivalent (precipitablewater) in depth from surface dew points. At least thisis shown to be the case for San Antonio in summer.REFERENCESReitan, C. H., 1963: Surface dew point and xvater vapor aloft. J. A ppl. Meteor., 2, 776-779.Bolsenga, S. J., 1965: The relationship between total atmospheric water vapor and surface dewpoint on a mean daily and hourly basis. J. Appl. Meteor., 4, 430-432.Smith, W. L., 1967: Note on the relationship between total pre cipitable water and surface dew point. J. Appl. Meteor., 5 . 726-727. Reply W. L. SmTaNational Environmental Satellite Center, ESSA, Suitland, Md.22 January 1968 In his comments on the relationship between totalprecipitable water and surface dew point, Schwarzillustrates the great difficulty of applying semi-emperical relationships derived from time and space averagedclimatological data to individual stations and synopticsituations. Due to the stochastic nature of weatherevents, empirical relations developed for meteorologicalpurposes can usually yield satisfactory results only forthe same time and space domain considered for thederivation of those relations. Thus, it is not surprisingthat the surface dew point is a poor predictor and thatthe X values derived from seasonal and latitudinalmean moisture profiles (Smith, 1967) are inadequatefor San Antonio, Tex., during various months of the>'ear. It was originally intended that these X values beused only to obtain estimates of the latitudinal averagevalues of total precipitable water from correspondinglatitudinal average values of surface dewpoint. Although it was stated that more characteristic Xvalues may be derived for individual stations, it isapparent from Schwarz's results that X should in manycases be allowed to vary with respect to time and thesynoptic situation. For those stations and conditionswhere a significant correlation does not exist betweensurface dew point and total precipitable water, thepredicting equation is, of course, useless. REFERENCESmith, W. L., 1967: Note on the relationship between total pre cipitable water and surface dew point. J. Appl. Meteor., 5, 726-727.