Laserfiche WebLink
2/rf <br />i0.6 _ --- - __ _ <br />;M 9 <br />u"~a+ in t.lr^ j-d-iv l,~.ricd orcc~~dinr .t ~.tt:-,n:.. .5.c~.. ~: ri ;-' :• ~ rr,~ ~i: <br />•i ~a~'~ AtdC-i i3 the ].owar limit of r:.~isture ~,_ _'~ -_ :. 1i:aiL :_' `, Fdrf'-11 <br />5~ T is the average for which the Cif of tabs '.1 at;ol••, and r+:::-ZII is ;`.^ <br />,n pP upper imlt of moisture or the lower liriL of S. T'h•:: E`i~- ^^ table 9. i <br />r'~ ~ were determined Uy means of r.aiYifall--`.:r.?f.' hlottings as c_::~ribed i:: <br />LI chapter 9 The same •~lotting~ ^~~rved '- ~cttin~ CPI fer '•C-T and <br />L <br />~~` <br />y <br />1` <br />• ,, <br />.~. r _ _.,. ~ .. _. _ <br />I,t}' I Pd•fC-III. That is, th•_ curves of fig•,rc _J.i w`._n s.i;,erini__sed on a <br />~~.' plotting also showed which curves lest °i: '.._ .::ghzs„ (t~:a-III) ano <br />lowest (AI.1C-I) thirds of t$e plotting. Tfie CIJ for high ara_oar mois- <br />~, tune levels ~.rere empirically velated to fire C?i of table 9._; the r~°- <br />~ sults are shown in cohimns 1, 2, and j of table 10.1, whic also gives <br />4or~ M values of Sand Ia for the CIJ in column 1. Tnz rainfall amo.mts on <br />berry which the selection of AtdC is based are given in table ~.2; th~~ dis- <br />~po:9 suasion in chapter 2 concerns the value of rainfall alone es a cri- <br />Pt terion for AIdC. Use of tables 4.2 and 1G.1 is demonstrates later in <br />~~ this chapter. In the section on courparisons of computed and actual <br />~ pj runoffs an example shows that for certain problers thz e:areme Ai;C <br />can be ignored and the average CId of table 9.1 alone applizd. <br />„~.I <br />D1 II R~,ATIODI OF Ia TO S. Equation 10.9 is base3 on the results shown in <br />figure 10.2 which is a plotting of Ia versus S for individual storms. <br />The data were derived from records of natural rainfall and runoff from <br />watersheds less than 10 acrzs in size. The largo amount of scatter in <br />the plotting is due mainly to errors ir. tine estimates of Ia. The magni- <br />bldes of S were estimated by plotting total storm rainfall and runoff <br />on figure 10.1, detel~ining the CIJ, and detetm.•ining the S :rom table <br />10.1. The magnitudes of Ia werz zstimated b;; taking the acc~iroulat.ed <br />rainiall from the begicming of a stox-m to tho tima when x•uno£f star.ecl. <br />E n•ors in S were due to determinations of average watershed rainfall <br />totals; these errors were very small. r.rx•ors in Ia wire d'.:e to one or <br />more of the following: (i) difficulty of determining the timz when rain- <br />fall began, because of storm travel and lac}; oY .instx•ccmentation, (ii) <br />difficulty of determining ttre time when runoff began, owing to the ef- <br />fects of x•ain on the measuring installations (channel runoff) and tv thz <br />lag of runoff from the watersheds, and (iii) impossibility of determin- <br />ing how much interception prior to runcff later made its way to the sail <br />surface and contributed to runoff; the signs and magnitudes of these er- <br />rors are not kno~m. Only enough points are _r.'_otted in figure 10.2 to <br />show the variability of the data. Thy line of relationship cute the <br />plotting into two equal numbers of points, and the slope of the line is <br />1:1 because the data do not indicate ot.hell+lse. A significant statisti- <br />cal correlation (chap. 18) between Ia and Scan bz made by adding more <br />points and increasing the "degrees of freedom.", but the standard error <br />of estimate will remain large owing to the deficiencies in the data. <br />L J <br />