|
<br />000388
<br />
<br />14
<br />
<br />IV - 4.3 - 4
<br />
<br />WARNING OR FORECASTING
<br />
<br />great. In the absence of any~ justifiable reason to the contrary and within the specifications provided in Figure IV -
<br />4.3 (I) an effort should be made to provide reasonably uniform density and thus preclude the possible occurrence
<br />of rel3tively large volumes of unreported precipitation.
<br />
<br />
<br />The design network for lIle study basin is shown in Figure IV - 4.3 (3), Examination of lIle figure may be
<br />considered to reveal two apparent deficiencies. In some cases the chance arrangement of forecast painls is so dense
<br />as to diminish their value as precipitation reporting stations and, in small isolated headwater catchments, one station
<br />in addition to the forecast point seems hardly adequate. It should be realized that the additional cost for reporting
<br />precipitation and other elements from a reporting river station is nominal and, considering the possibility of mal.
<br />function of one station, it is probably advantageous to collect precipitation from aJl river stations unless they are
<br />within a few km of each other. In those cases where an appreciable portion of the drainage divide of a catchment
<br />constitutes aho a' boundary of the network (see catchments of 640, 275 and 583 sq. km. along the northwestern
<br />edge of basin in Figure IV - 4,3 (2), a design function such as in Figure IV - 4,3 (I) will provide inadequate denSity
<br />and stations should be added. Usually, however, there will be reports from adjacent basins and rhis will not be neces.
<br />sary.
<br />
<br />Other considerations summary and conclusion
<br />
<br />One rather simplified approach to the design of a reporting network has been presented which was developed
<br />and applied throughout the country several years ago largely for planning purposes. The results have also been most
<br />helpful in evaluating the relative adequacy of the existing network in different regions of the counuy. 11 is believed
<br />that the a.pproach has some merit for such purposes, although a more rigorous design study is envisioned to installa.
<br />tion of any major segment of an adequate automated network.
<br />
<br />~J""
<br />. "..\'f,
<br />"~R~iJj
<br />
<br />The illustration presented neglects any consideration of the ex..isting network and communications facilities as
<br />well as the accessibility of station sites and the fact that funding rarely pennits rapid increases in network density
<br />on a nation3J scale. The nature and costs of installation and moving of automated stations are such that any incre.
<br />ment of funding should be spent in upgrading the network in a selected area to design density.
<br />
<br />Without the advantages of automation, availability of observers and communications facilities arc overriding
<br />considerations in the placement of stations, These factors are not particularly relevant in the case of automated nct.
<br />works, but accessibility and availability of power must be taken into account.
<br />
<br />REFERENCES
<br />
<br />I. L'tngbein, W. B. Error in computation of mean (IreaJ rainfall, US Geol. Survey, Aug_ 1946 (processed).
<br />
<br />2_ Light, Phillip. The reliability of are::tI rJinfall determinations, Hydrometcorological Rep. S, US Weather Bur..
<br />Corps of Engineers. 1947,
<br />
<br />3. linsley, R. K. and Kohler, M. A_ Variations. in storm rainfall ovu small areas, Trans_ Am. Geophys. Union,
<br />vol. 32. pp, 245-250, 1951,
<br />
<br />4, Huff. F, A, and Neill, J, C. R,infall relations on small areas in Illinois. III, State Water Surv, Bull. 44,
<br />61 pp, 1957,
<br />
<br />5,
<br />
<br />McGuinness, J. L. Accuracy of estimating watershed mean rainfall, J. Geophys. Res., vol. 68, 4763-4767,1963,
<br />
<br />6,
<br />
<br />Linsley, R, K,. Kohler. M, A, and Paulhus, J, L. H, Hydrology for Engineers, p, 31. McGraw-Hili, New York,
<br />1958,
<br />
<br />(
<br />~..
<br />
|