Laserfiche WebLink
<br />. <br /> <br />. <br /> <br />identified. The regression line will be plotted, along with all data <br />points,:using the format in figure 2. Significance of the linear trend <br />Mill beievaluated by a student-t test of the regression slope parameter, <br />~ Mith th~ null hypothesis that the slope is zero. <br />~ <br />~ Step 4. - Investigation of concentration-streamflow relationships: <br />o <br />Any of lhe six solutes exhibiting significant trends in concentration, <br />load. or, mass fraction will be subjected to additional analyses to deter- <br />mine whither the long-term concentration-streamflow relationship has a <br />signifiqant time component. First, concentration will be plotted as a <br />functioq of streamflow, using the complete data set, to determine the <br />nature qf the relationship. Simple mathematical functions (e.g., power, <br />exponen~ial, logarithmic, quadratic) will be fit to the data. Again, <br />values ~reater than two standard deviations from the model predictions <br />Mill be :identified as outliers. <br /> <br />Paramet~rs of the selected model will then be modified to include a time <br />term, arid the resulting model will be fitted empirically to the complete <br />data seti. The significance of the coefficient on time will be determined <br />by a stUdent-t test. Each parameter will be evaluated individually. For <br />example~ if the postulated relationship is <br /> <br />Bl <br />eim = BO Qm <br /> <br />.'the follpwing two models may be applied: . <br />(Bl + B2t) <br />eim = BO Qm <br /> <br />B2 <br />eim = (BO + BIt) Qm <br /> <br />where Bniare empirical parameters and t is time in months since the <br />beginninw of the period of record. <br /> <br />Step 5..- Elim~nation of outliers: <br /> <br />Outlyingidata points listed in steps 1 through 4 will be reexamined to <br />determin, the possibility they result from bad or anomalous data. Any <br />values sq identified will be eliminated from the data set. Steps 3,and 4 <br />will be repeated using the revised data. <br /> <br />Interpretation of Identified Trends <br /> <br />Long-term trrends or step changes in streamflow-concentration, solute load, or <br />solute mass 'fraction relationships will be traced from downstream to upstream <br />gaging stations to isolate their potential source. Possible causes which, <br />Mill be evaliuated include changes in data collection, analysis, or reporting <br />procedures, land initiation of water development projects such as irrigation <br />or impoundm~nt. Mass balanceswill be computed between stations to determine <br />salt loadin~ as a function of time. <br /> <br />'1. <br />;~ <br />, <br /> <br />;] <br />'.'1, <br />'I <br />" <br /> <br />8 <br /> <br />,~ <br />I <br />~ <br />~!i <br />, <br /> <br />. <br /> <br />,.. '. "'_T"""'!!l"~'",..,..,_. ~.'~_ ~4~.~'"'"''.,-: --;, ;-~;_P' ~; --\f "O:'';~" '< , <br /> <br />,:","':",,',';'" <br /> <br />':?'":'-'.'-' ,'4 <br /> <br />0' _..J~~'"'__~", '," ,- .i.._,..", <br />