cooler in the surface layers than the air normally present over the tropical oceans in summer. This also holds true in air with a trade inversion and upper dry layer present. Surface ship temperature and dew point reports, along with   upper   air   data,   are   extremely   valuable   in determining whether the surface layers are truly tropical or whether they contain old polar air not yet completely transformed. THE DYNAMICS OF TROPICAL CYCLONES Dynamically, storms (existing and potential) are subject to influences from the surrounding areas. In the Tropics,  we  usually  encounter  two  layers  in  the troposphere,  and  these  levels  have  very  different characteristics.  Most  of  the  time,  a  steady  trade  wind blows in the lower levels (surface to 500 hPa), while a succession of large cyclonic and anticyclonic vortices are present in the upper troposphere in the 400- to 150-hPa  strata.  In  some  regions,  such  as  the  area between  the  Mariana  Islands  and  the  South  China  Sea during midsummer, intense eddy activity also takes place near the surface. It is not possible to deduce from charts drawn in the lower layers whet is taking place above 500 hPa. Therefore, it is necessray to keep track of events in both layers. Aside from the surface chart, the 700-hPa level chart  is  very  helpful  in  determining  low-level  flow patterns. The 200-hPa level is representative of the upper layer, whereas the 500-hPa level, often located in the transition zone, is of much less use in tropical than in  extratropical  forecasting. However, forecast considerations should not be limited to the two layers in the Tropics. Middle latitude weather also influences and shares in the control of weather changes in low latitudes. The position and movement of troughs and ridges in the westerlies affect both  the  formation  and  motion  of  tropical  storms. Therefore, middle latitude analysis is important. Since forecasts generally run from 1 to 3 days, there should also be a hemispheric analysis. The  climatological  approach  to  the  forecast problem  should  also  be  taken  into  consideration.  It  is obvious that a forecaster must be familiar with regional and  seasonal  changes;  for  instance,  areas  of  frequent tropical  cyclone  formation,  mean  storm  tracks,  the scattering of individual tracks about the mean, and the month to month variations of all of the above. Mean tracks and other data on Pacific storms are available in several publications. These maps and charts reveal that the   “climatological”   approach   gives   some   useful information, but it cannot be relied on in anyone specific area,  or  in  the  case  of  any  particular  storm  to  the exclusion  of  synoptic  indications  for  the  forecast.  These probability   considerations   are   used   mainly   for long-range  planning  and  when  data  are  unavailable. Therefore, we reach the conclusion that climatological information should be treated as a weighing factor to be included  after  evaluating  synoptic  data. FORMATION A full-blown typhoon/hurricane cannot be forecast as such when there are no indications of any type of irregularity  on  the  charts  or  aids  used  in  the  tropical analysis.  Only  when  the  incipient  stage  frost  appears among the data can we begin to think in terms of an actual typhoon/hurricane, and often not even then. For the  most  part,  the  forecast  evolution  from  “area  of disturbed weather” to typhoon/hurricane is a matter of step-by-step  progression.  Assume  the  term  formation means formation of a “potential” typhoon/hurricane. Empirical rules or checks have evolved through the years that enjoy a measure of reliability to warrant their use,  and  of  course,  the  more  signs  that  point  to formation, the more likely formation will occur. Synoptic Conditions Favorable for Development The following list contains conditions favorable for development  of  tropical  cyclones.  In  this  list,  no attempt is made to separate the surface from the upper air indications, as the two are most often occurring simultaneously and are interrelated. l Marked cyclonic turning in the wind field. .  A  shift  in  the  low-level  wind  direction  when easterlies  are  normally  present.  Wind-speeds  generally 10 knots or more. . Greater than normal cloudiness, rainfall, and pressure falls. Cloudiness that increases in vertical as well as horizontal extent. .  Moderate  to  strong  outflow  aloft. l Sea level pressures lower than normal. The value is dependent upon the region analyzed; however, a deviation of greater than 3 hPa is the normal criterion. .  Easterly  wind  speeds  25  percent  or  more  above normal in a limited area, especially when the flow is Cyclonic. 11-2