NSWC/WOL/TR 75-3 5 



kinds of natural waters. The most important requirements for a 

 satisfactory sample are that it be both valid and representative. 

 For a sample to be valid, it "has to be one which has been collected 

 by a process of random selection. Any method of sampling that 

 sacrifices random selection will impair statistical evaluation of 

 the analytical data. 



A satisfactory sample is not only randomly drawn, but also is 

 representative. This means that the composition of the sample should 

 be identical to that of the water from which it was collected; the 

 collected sample should have the same physiochemical characteristics 

 as the sampled water at the time and site of sampling. 



Planning for a sampling program should be guided by the overall 

 objectives of analysis. Major factors of concern for any sampling 

 program are: (a) frequency of sample collection, (b) total number 

 of samples, (c) size of each sample, (d) sites of sample collection, 

 (e) method of sample collection, (f) data to be collected with each 

 sample, and (g) transportation and care of samples prior to analysis. 



Frequency of sampling will depend to a large extent upon the 

 frequency of variations in composition of the water to be sampled. 

 There are two principle types of sampling procedures commonly used 

 for analysis of natural waters. The first type is that which yields 

 instantaneous spot or grab samples, while the second type yields 

 integrated continuous or composite samples. A grab sample is a 

 discrete portion of water taken at a given time; a series of grab 

 samples reflects variations in constituents over a period of time. 

 The size of such individual samples will depend on the objectives 

 and methods of analysis, and on the required accuracy. The total 

 number of grab samples should satisfy the statistical requirements 

 of the sampling program. 



Composite samples are useful for determining average conditions. 

 A composite sample is essentially a weighted series of grab samples, 

 the volume of each being proportional to the rate of flow of the 

 water at the time and site of sample collection. Samples may be 

 obtained over any time period such as 4, 8, or 24 hours, depending 

 the purposes of analysis. 



Selection of sampling sites should be made with great care. 

 A field survey is often useful in planning for site selection. 

 Special consideration should be given to sources of discharge, 

 dilution by tributaries, and changes in surrounding topography. 



14 



