will be noted that the south-east coast of the islands has a definite convex 

 curvature. It has been shown that low pressure systems approaching a 

 convex coast line tend to weaken. 



Extrapolation, climatology, and surface synoptic steering were invest- 

 igated as methods of forecasting the direction and speed of movement of 

 typhoons. Extrapolation was found to be useful, if applied to good data, 

 especially either in the early or late parts of the season. Extrapolation 

 was a very poor method of forecasting in those cases in which the present 

 and past position of the typhoon were inaccurate, and also poor during 

 July, xA-ugust, and September, when the tracks are most erratic. Another 

 difficulty in using extrapolation in forecasting the rate of movement was 

 that many of the typhoons (notably Gwen, Kathleen, and Rosalind) did 

 not accelerate upon recurving. 



Climatology was especially useful in 1947. In many instances climat- 

 ology was the only aid available, and late in the season during November 

 and December the typhoons followed the tracks of climatological expect- 

 ancy well. Of course, some difficulties were encountered in using the 

 climatological data available. Tracks shown in different publications do 

 not agree. Those shown in H.O. 219, " Climatology, Asia Air Station,'" 

 appear incorrect, especially the average tracks shown for November, 

 December, and January. x\lthough climatology given in " Typhoon 

 Tracks Supplement to Weather Guides for Long Range Planning," pre- 

 pared by the War Advisory Council on Meteorology, U.S. Weather 

 Bureau, March, 1944, covers a period of approximately twenty years, it 

 apparently does not include all the possible variations. 



In conclusion, some improvements that could be made to the typhoon 

 warning service in the north Pacific should be pointed out. In order to 

 devise better methods of forecasting the direction and movement of typ- 

 hoons detailed study of future typhoons is necessary. Post analysis 

 reports are necessary, similar to that made by Kidd and Reed and this 

 report. The 2143d Air Weather Wing has directed that the Harmon 

 Weather Central prepare an annual report of this type. But these re- 

 ports can only do a portion of the task. Many of these reports lack 

 valuable data from the P.I., China, and Indo-China. A first step would 

 be to establish an effective exchange among all agencies of new data for 

 post-analysis purposes. 



Discussion 



Mr. Simpson pointed out that the reasons for certain anomalous 

 typhoon tracks could sometimes be found in the filling or deepening of 

 upper level troughs, perhaps by the mechanism of wave transfer of 

 energy given prominence by Namais. Mr. Gringorten pointed out that 

 there were difficulties in using the simple diagrams in which the direction 

 of motion of a typhoon is obtained by vector addition of a cyclonic 

 current and a general carrying current. 



Mr. Mordy also pointed out that this idea appeared to involve inflow 

 to the south of the typhoon and outflow to the north, which was contrary 

 to experience. 



In connection with Captain Woodall's statement that convection 

 should not be over-emphasized in the formation of tropical storms, most 

 of which formed over the open sea and not over islands, Mr. Green 

 remarked that in his experience convention was always active in the 

 formation of the storms, but was often provided b\' tropical convergence 

 zones and not by islands. 



33 



2 — Pac. Congress 



