Rises of this nature on the open coast can also affect levels in bays and 

 estuaries. 



There are other causes for departures of the water levels from normal in 

 semienclosed and enclosed basins, such as the effects of evaporation and 

 rainfall. Generally, rainfall plays a more dominant, role since these basins 

 are affected by direct rainfall and can be greatly affected by rainfall runoff 

 from rivers. The initial rise caused by rainfall is due to rains preceding 

 the storm; rains during the passage of a storm have a time-dependent effect on 

 the change in water level. 



e. St orm Surge Prediction . The design of coastal engineering works is 

 usually based on a life expectancy for the project and on the degree of 

 protection the project is expected to provide. This design requires that the 

 design storm have a specified frequency of occurrence for the particular 

 area. An estimate of the frequency of occurrence of a particular storm surge 

 is required. One method of making this estimate is to use frequency curves 

 developed from statistical analyses of historical water level data. Table 3- 

 6, based on National Ocean Service tide gage records, indicates observed 

 extreme storm surge water levels including wave setup. The water levels are 

 those actually recorded at the various tide stations and do not necessarily 

 reflect the extreme water levels that may have occurred near the gages. 

 Values in this table may differ from gage station values because of cor- 

 rections for seasonal and secular anomalies. The frequency of occurrence for 

 the highest and lowest water levels may be estimated by noting the length of 

 time over which observations were made. The average yearly highest water 

 level is an average of the highest water level from each year during the 

 period of observation. Extreme water levels are rarely recorded by water 

 level gages, partly because the gages tend to become inoperative with 

 extremely high waves and partly because the peak storm surge often occurs 

 between tide gage stations. Poststorm surveys showed water levels, the result 

 of Hurricane Camille in August 1969, in excess of 6 meters (20 feet) MSL over 

 many miles of the open gulf coast, with a peak value of 7.3 meters (24 feet) 

 MSL near Pass Christian, Mississippi. High water levels in excess of 3.7 

 meters (12 feet) MSL on the open coast and 6 meters (20 feet) within bays were 

 recorded along the Texas coast as the result of Hurricane Carla (September, 

 1961). Water levels above 4 meters (13 feet) MSL were recorded in the Florida 

 Keys during Hurricane Donna (1960). 



Accumulation of data over many years in some areas, such as regions near 

 the North Sea, has led to relatively accurate empirical techniques of storm 

 surge prediction for some locations. However, these empirical methods are not 

 applicable to other locations. In general, not enough storm surge obser- 

 vations are available in the United States to make accurate predictions of 

 storm surge. Therefore, it has been general practice to use hypothetical 

 design storms and to estimate the storm-induced surge by physical or mathe- 

 matical models. Mathematical models are usually used for predicting storm 

 surge, since it is difficult to represent some of the storm surge generating 

 processes (such as the direct wind effects and Coriolis effects) in physical 

 laboratory models. 



3-115 



