Laserfiche WebLink
<br />-r------__,.t.1.__. _ ...______./ <br /> <br /> <br />. J;, <br /> <br />TN RE:PLY REFER TO: <br /> <br />United States Department of lhe Interior <br /> <br />GEOLOGICAL SURVEY <br />Box 25046 <br />Denver Federal Center <br />Denver. Colorado 80225 <br />Ma il Stop #415 <br /> <br />Water Resources Division <br />Colorado District <br /> <br />Apri I 21, 1977 <br /> <br />~ -'I:';~J '; r. -, , <br />inl,". I, J, i ''', <br />"" If >.-r" __" l 'I, <br />11117 ..L." ~. . <br />I' <br />" Mr.IY' 'I' <br />'I I,. I <br />"i", <br />.\.._."L. <br /> <br />., I <br />. ._1 <br />, Iii <br />" <br /> <br />Wen-Sheng Liang <br />Camp, Dresser & McKee, Inc <br />1660 South Albion Street <br />Denver, CO 80222 <br /> <br />i <br />L , <br /> <br />': .. <br /> <br />Dear Sir: <br /> <br />':>~-NSt-_~; .':,'" <br /> <br />!, " <br /> <br />I have finally found time to review your ptoposed nood discharcles <br />for the HUD Study of San Miguel County. I have several cOITments <br />as follows: <br /> <br />Q-~,"O\..UT/O'; ~('1 <br />~ ~ <br />23 r~ <br />it. .;~ <br />~ ~. <br />If,.. "'~ <br />'?'76_191X> (~ <br /> <br />1) Your regression relations are flatter than the comparable ones <br />in CWCB Technical Manual No. 1. Ihis is surprising because I <br />intuitively feel that separating the rainfall and snowmelt <br />components should steepen the regression relations. Several <br />stations used by you may in part I?xplain the problem. <br /> <br />2) Gaging Station No. 09174500 Cottonwood Creek near Nucla p'lots <br />extremely low on your relation. We did not use the station <br />for TM #1 because of two extremely low peaks (they probab'ly <br />pass the low outlier test specifil2c1 in wRC Bullet'in No. 17 but <br />they sure cause an abnonnal negative skew). I wou'ld suggest <br />deleting this station. <br /> <br />3) The two large stations on the San Miguel River tend to aml)'lify <br />the flat regression slopes because they plot very 'low on your <br />relation. Conversely, the two large stations on the Dolores <br />River plot high on your relation. Maybe putting the latter <br />two into your regression would y'ielcl a b~tter solution. <br /> <br />I strongly feel that including large stations and small stations <br />into an overall regression also yield flatter slopes. For <br />example, leaving out the San Miguel River and Cottonwood Creek <br />data, your relation would2probably be drastically steepened <br />in the DA range 10-100 mi . Valley storage on the larger <br />basins probably causes much of this effect. <br /> <br />4) <br />