AEROLOGY OF TROPICAL STORMS 
wind data are often missing completely. Gradually, 
rawins should replace pibals everywhere. 
The value of the raob in tropical analysis has been 
much discussed. This writer, along with others, used 
to think that they were useless. Horizontal gradients 
of pressure and temperature in the tropics are very 
small. At 300 mb, the height gradient may be as low as 
200 ft per 10 degrees of latitude. A small error in 
sounding evaluation will produce an error of this magni- 
tude at 300 mb. Moreover, local horizontal tempera- 
ture gradients can be as large as the synoptic gradients. 
A single observation with such errors can grossly distort 
a whole map pattern. 
In spite of these difficulties, raob analysis has turned 
out to be valuable when we are concerned with reports 
from stations where the quality of observations is high, 
such as San Juan in Puerto Rico, Swan Island, and 
Honolulu. Time sequences, especially when considered 
together with wind data, are reasonable. Obviously 
incorrect soundings are rare at such stations and stand 
out very clearly, especially in the case of nighttime 
observations. The soundings taken during the day give 
more erratic results and are not so useful. It is suggested 
that nighttime soundings be continued and expanded. 
Methods of Analysis. Area to be Analyzed. As already 
noted, recent years have brought to light the strong 
influence that the upper layers exert on surface con- 
ditions. They have also revealed the large extent of 
interaction across the subtropical ridge. Thus tropical 
analysis should be extended to high levels and high 
latitudes. Writers, however, disagree on the usefulness 
of eross-equator analysis for short-term hurricane fore- 
casting. Some, for example, state that West Pacific 
and Indian Ocean typhoons form following an intensi- 
fication of flow across the equator. Others maintain 
that such an increase occurs after and in consequence 
of typhoon development. This writer has seen instances 
of the second type of situation but would not maintain 
that this is always the case. Experimental analysis 
across the equator is obviously useful. But as yet we 
cannot say that its value for short-term forecasting has 
been proven. 
Contour Analysis. As in higher latitudes, the mterest 
of the analyst centers on the fields of mass and motion. 
Since tropical storms decrease in intensity upward, 
surface analysis is a logical tool in hurricane situations. 
Outside of storm areas, Riehl and Schacht [32] have 
proposed adoption of the 850-mb (5000-ft) level as the 
basic level in order to eliminate many of the local 
features contained in surface reports. 
Upper levels in regular or experimental use are 700, 
500, 300, and 200 mb. The 700-mb level is situated in 
the center of the lower trade stream and reflects the 
disturbances in this current very well. At 500 mb we 
are in the zone of transition from trade-wind to upper- 
vortex regime. Winds are light and variable and height 
gradients are erratic. This surface often is not suitable 
for hurricane work, especially in the Pacific. Even the 
300-mb level is not very satisfactory, although it serves 
well for many purposes of general tropical forecasting. 
Over the western portions of the oceans in summer, the 
903 
high tropospheric vortices reach their greatest intensity 
near 200-150 mb. Analysis of the 200-mb level is 
essential for hurricane forecasting. 
Because of the uncertainties involved in drawing 
contours in the tropics, described earlier, differential 
analysis and time sections must supplement the con- 
tour fields at all times. There is also the question of the 
pressure-wind relation. Opinions differ widely as to the 
latitude at which the gradient-wind relation becomes 
useless. It has been noted that aircraft dispatched with 
the ‘‘pressure-pattern” method [1] have arrived at their 
destination with considerable exactness in low latitudes. 
Some authors wish to apply the thermal-wind equation 
even at latitude 5°. Of course, there must be variations 
with synoptic conditions. In a strongly disturbed region, 
cross-isobar flow will be pronounced. It is of obvious 
importance to determine the true nongeostrophic flow. 
This could be done by using the pressure-pattern method 
to dispatch aircraft at different levels in selected situa- 
tions and calculating the drift of the aircraft and the 
error between the computed and actual flight time. 
Until this is carried out, it is hazardous to make assump- 
tions about the size of the angle between streamlines 
and contours when drawing upper-level charts. 
Streamline Analysis. Because of the weak contour 
gradients that prevail especially at low levels outside 
of storm areas, streamline analysis can brmg out many 
details of the field of motion. Some analysts, including 
the writer, at times have replaced contour and isobaric 
analysis entirely with qualitative streamline methods. 
The lines are drawn parallel to the wind direction and 
qualitative allowance for convergence and divergence 
is made by starting and ending some streamlines. Al- 
though this technique leads to clear and illustrative- 
looking charts, it has a distinet drawback: the charts 
are not suited for computations. An alternative is the 
construction of exact streamline and convergence charts. 
There is no doubt that exact streamline fields are one 
of the best potential tools for future research, especially 
for research concerning the structure of hurricanes. 
Stability Charts. Since the thermodynamic hypothesis 
of tropical storms is the oldest, the literature carries 
many descriptions of vertical stability conditions in 
and near hurricanes. The trade-wind inversion dis- 
appears in storms. It has been a favorite argument to 
say that moisture accumulates under an inversion over 
the oceans. Then, as the inversion breaks down through 
heat accumulation below, or for some other reason, 
large quantities of water vapor suddenly become avail- 
able for upward transport and condensation. The release 
of this latent heat provides an unbalanced energy source 
and can initiate storms. 
This writer has applied this argument to the problem 
of driving mechanisms for the general circulation in 
low latitudes [29]. Local application to incipient hurri- 
canes, however, has not proved fruitful. Storms often 
form in an atmosphere that has been nearly moist- 
adiabatic for a long time. We know that the steepest 
lapse rates occur in subsiding dry air above the trade 
inversion. Hxcept for analysis of the inversion itself, 
the writer believes that stability charts are not likely 
