. Storms developing on the eastern edge of polar troughs initially move up the trough. .  Pacific  storms  forming  on  equatorial  shearlines usually are most active in the southern and southwestern portions  at  the  start  and  they  move  slowly  northeast. After 1 to 2 days, the influence of the trades becomes dominant and the storms turn back toward a direction ranging from west to north. (This appears to be true mainly for cyclones deepening near or west of 130°E.) l If a storm is discovered without information as to its past movement and the upper air current, it is best to start it along the climatological mean track and then secure the data necessary to determine the steering current. Extrapolation At  present,  the  most  practical  prognostic  technique used   by   the   tropical   meteorologist   consists   of extrapolating  the  past  movement  of  the  synoptic features on his or her chart into the future. The past track of a cyclone represents the integrated effects of the steering  forces  acting  upon  it.  Accelerations  and changes in course are the results of the changes in these steering forces. The  effect  of  these  forces  can  be examined  and  extrapolated  directly  from  the  past position  of  the  cyclone. Before applying the extrapolation technique, the forecaster  must  attempt  to  smooth  out  the  minor irregularities in the past track. The first step in the use of  extrapolation  is  to  determine  the  mean  direction  and speed  of  the  cyclonic  center  between  each  two  known positions. The next step is to determine the rate of change indirection and speed between successive pairs of fixes. The forecast position thus determined from extrapolation of movement should be smoothed out for minor irregularities and compared to the applicable climatological tracks of cyclones in the area. If large differences exist, the forecast should be completely reexamined.  The  climatological  tracks  should  receive less weight for short-term forecasts of 6 to 12 hours and more weight for forecasts in excess of 24 hours. For   short-term   forecasts   (6   to   12   hours), extrapolation is as reliable as any known method for movement. Steering The  movement  of  a  tropical  cyclone  is  determined to a large extent by the direction and speed of the basic current in which it is embedded. This concept appears to work well as long as the cyclone remains small and remains in a deep broad current. By the time a tropical cyclone  has  reached  hurricane  intensity,  these conditions  seldom  exist.  It  then  becomes  necessary  to integrate the winds at all levels through which the cyclone extends and in all quadrants of the storm to determine  the  effective  steering  current.  Although  the principle is not fully understand and has not yet received universal  acceptance  as  a  valid  rule,  it  does  have practical  applications.  For  example,  changes  in  winds atone or more levels in the area surrounding the cyclone can sometimes be anticipated, and in such cases, a qualitative  estimate  of  the  resulting  change  in  the movement of the cyclone can be made. Conversely, when it appears likely that none of the winds in the vicinity  of  the  cyclone  will  change  appreciably  during the forecast period, no change in the direction and speed of movement should be anticipated. Further,   this   rule   applies   to   situations   of nonrecurvature and only some cases of recurvature. It is difficult to determine the steering current, since the observed  winds  represent  the  combined  effects  of  the basic current and disturbances. As most storms extend into the high troposphere, it is better to calculate a “steering  layer”  than  a  steering  level,  since  presumably the  wind  throughout  most  of  the  troposphere  influences the  storm  movement.  One  writer  recommends  an integration  of  the  mean  flow  between  the  surface  and 300 hPa, over a band 8° in latitude centered over the storm. Another writer indicates that for moderate and intense  storms  the  best  hurricane  steering  winds  would be found in the layer between 500 and 200 hPa and averaged  over  a  ring  extending  from  2°  to  6°  latitude from the storm center. STREAMLINE  ANALYSIS  AND  VECTOR AVERAGES.—  Other practical applications of the steering concept to short-range tropical cyclone motion use differing approaches in attempting to measure the basic  current. One  is  by  streamline  analysis  of successive levels to find a height at which the vertical circulation diminishes to a point such that the winds are supposedly  representative  of  the  undisturbed  flow. Another   method   is   to   take   vector   averages   of reconnaissance winds near the zone of strongest winds in the storm. USE OF OBSERVED WINDS ALOFT.— When sufficient data are available, it has been found that the use  of  streamline  analysis  of  successive  levels  usually gives  valuable  indications  of  tropical  storm  movement for as much as 24 hours in advance. However, since wind observations are usually scarce in the vicinity of a 11-5