Laserfiche WebLink
<br />Q 000749 <br /> <br />r:C') <br />~. <br /> <br />-4- <br /> <br />I <br />PROBABILITY ANALYSIS OF STREAMFLOW DATA <br /> <br />Probability a~is is a statistical and mathematical <br />process for determining the probable occurence of a given pheni:llll8non <br />based on observation of the previous occurence or occurences of that <br />phenomenon~ Several types of analysis of probability have been <br />developed which. are applicable to various types of phenomena. The <br />most COIlllllOn type of analysis is the so-called "normal frequency <br />distribution" type. This is mainly applioable to events whose <br />ooourence is not related to t1me. The presenoe of the t1me factor, <br />as in a series of annual streamf'lows, usually requires the 8ll8lysis <br />of the possibility of sequential oorrelations or linkages of two or <br />more years within the total series. The Colorado River data will be <br />tested for randCllllnllss and for J i "lrtoge. <br /> <br />During the course of the probability analysis desoribed <br />above we would evAmine examples of the application of probability <br />analysis to streamflow data and also eYAminA the literature pertain- <br />iDg to this problEIDI. We would prepare a. digest of this literature <br />for the use of the Board and also prepare special reports to the <br />Board on significant applioations of the analysis. We call yOur <br />attention, for example, to the test1molW of Dr. Luna B. Leopold, <br />Chief Hydraulics Engineer, USGS, in Arizona versus California. In <br />his test1moll1 Dr. Leopold presents an analysiS based on the use of <br />the normal frequency distribution curve. We believe that it is <br />very 1mportant for the staff of the Board to have at its disposal an <br />evaluation of the validity of this approach as well as a comparison <br />between it and alternative techniques of probability ana4>'sis. <br /> <br />FACILITIES <br /> <br />The facilities of the Bureau of Economic Researoh will <br />be available for the oar:rytng out of this proposal. <br /> <br />PERSONNEL <br /> <br />The probability ana4>'sis will be made by Dr. Morris E. <br />Garnsey and Dr. Leslie Fishman, both professors in the Department of <br />Economics of the University of Colorado. Dr. Garnsey is a speoialist <br />in Regional Economics. Dr. FishmBn is a statistician. They will <br />oonsult with Mr. Crow and Mr. Macdonald. <br /> <br />Dr. Garnsey will supervise all the phases of the over-al1 <br />proposal and ooordinate the analyses of streamflow, preoipitation <br />and atmospherio ohanges. <br />