Laserfiche WebLink
<br />14,000 <br /> <br />o <br />z <br />o <br />() <br />~ 12.000 <br />a: <br />W <br />"- <br />Iii <br />~ 10,000 <br />() <br />iii <br />:J <br />() <br />~ 8,000 <br />W <br />(!) <br />a: <br /><( <br />J: <br />~ 6,000 <br />i5 <br />::; <br />Lli <br />a: <br />t:i 4,000 <br />Z <br />Lli <br />::; <br />-' <br />~ 2,000 <br />Z <br />Z <br /><( <br /> <br />~ <br />1\ <br />I \ <br />I \ <br />I \ <br />~ h. <br />I - -q <br />I \ <br />I \ <br />1 \ <br />I \ 0 <br />Vl., 1 \ 1 <br />--~--~~--------r~---i- ---~----------f- <br />\ / ~~ / \ ~ \ I <br />\ I \ j \ I \ I <br />\ I \ (f) \ I \ I <br />(j, \ I \ I ~ <br />\I <br />o <br /> <br /> <br />__, !. r C' <br />':;"~ V -' <br /> <br />l3-f] <br /> <br />STATION 09095500 <br /> <br /> <br />"'1>." <br /> <br />...,' <br /> <br />'A. ..t>...Ji <br />"'6."Ii" <br /> <br />o <br />1968 1970 1972 1974 1976 1978 1980 1982 1984 1986 1988 1990 1992 1994 <br />WATER YEAR <br /> <br />I!...[;. STATION 09152500 <br /> <br />G-O <br /> <br />STATION 09163500 <br /> <br />Figure 3. Annual mean stream discharge for gaging stations 09095500 Colorado River near Cameo, <br />09152500 Gunnison River near Grand Junction, and 09163500 Colorado River near the Colorado-Utah <br />State line, water years 1970-93. <br /> <br />Horizontal lines are lhe mean annual <br />stream discharges, water years 1970-93 <br /> <br />.... ..6."4:'.:.:".:.l<.~:.~.. <br /> <br />...... . <br />Q '. <br /> <br />METHODS OF TREND ANALYSIS <br /> <br />Trend analysis can be used to determine if <br />stream-water quality has changed over time. Two <br />general types of trend teslS, monotonic and step-trend <br />tests, were used to examine dissolved-solids and major <br />constituent data associated with the salinity-control <br />projects. A monotonic trend means that the water- <br />quality variable of interest has changed over time. <br />A monotonic trend test will not specify if the change <br />occurred continuously, linearly, or in abrupt or discrete <br />steps (Hirsch and others, 1991). Slep-trend tests are <br />used instead of monotonic tests where a known event <br />or change occurred at a specific time in a watershed that <br />could have significantly altered constituent concentra- <br />tions or loads. In such cases, the data can be divided <br />into "before" and "after" periods relative to the known <br />event An example of such a known event is the well <br />plugging for the Meeker Dome Unit Another case for <br />use of step-trend tests is when the water-quality record <br />is broken by a relatively long time gap. Conversely, <br />step-trend tests would not be used if water-quality <br /> <br />changes were incremental over a rather long time <br />period. For a long-term effect, such as the salinity- <br />control projects in the Grand Valley and lower <br />Gunnison River Basin, monotonic trend tests are more <br />appropriate for trend analysis. <br /> <br />Monotonic Trend Analysis Using the <br />Seasonal Kendall Test <br /> <br />Monotonic trends on periodic concentration data <br />and monthly dissolved-solids loads were examined <br />using a computerized procedure developed by the <br />USGS called EStimate TREND (ESTREND) (Schertz <br />and others, 199]). The program is designed to investi- <br />gate trends in water-quality data that often have non- <br />normal distributions, seasonality, outliers, missing <br />data, or have censored data (values reported as less <br />than). The program is written in FORTRAN language <br />for the USGS Prime minicomputer system (Schertz and <br />others, ]991). <br /> <br />10 Trend Analysis 01 Selected Water-auality Data Associated With Sallnity-Control Projects In the Grand Valley, <br />In the Lower Gunnison River Basin, and at Meeker Dome, Western Colorado <br />