base his estimate on the conditions he anticipates from his examin- 

 ation of the weather maps. The following or the opposing winds can 

 be estimated in a similar manner. 



In order to arrive at an estimate of the rapidity with which 

 swell may die out it is advisable to split the fetch into several 

 parts and compute the swell from each. 



In middle latitudes a sequence of low-pressure systems, that 

 is, a sequence of generating areas, often travels across the oceans. 

 It is recommended that the swell which is forecast from each genera- 

 ting area be plotted on graph paper, using height of swell and time 

 of arrival as coordinates. Observed values should be entered on the 

 same graph in order to test the accuracy of the forecasts. 



In carrying out the forecasting it may be found that several 

 wave trains arrive at approximately the same time; in this case the 

 resulting swell will be complicated because of interference . The 

 greatest v/ave heights may eoual the sum of the heights in the indiv- 

 idual wave trains but the average height mil be that characteristic 

 of the train having the highest waves. It appears probable that 

 v/ith experience the complexity of the expected swell can be forecast. 



The general procedure which has been outlined should be modified 

 according to the type of weather maps which are available and accord- 

 ing to the experience of the forecaster. However, it should be em- 

 phasized that the c ontinuity of the processes must be borne in mind . 



Example 



Forecast of swell for Casablanca and vicinity, 

 Northwest coast of Africa, November 7, 1931- 



51 



