WMOD00575 - Laserfiche WebLink

Home Browse Search

DATA USED FOR THIS STUDY Chapter IV The data used in this study are winter and spring precipitation and spring runoff. They have to be col- lected in a certain order and have to satisfy specific criteria. These conditions are discussed in this chap- ter. 1. Precipitation and runoff in the Upper Colorado Basin. a. Precipitation records. According to the United States Weather Bureau's "Substation History" (26-30), about 400 stations are found in the Upper Colorado River Basin, including stations with records of storage gage and stations not now in operation. For 312 of these stations, monthly precipitation data were collected from the following sources and recorded on magnetic tapes. (1) The United States Weather Bureau, "Climatological Data" [31,35] (2) The United States Weather Bureau, "Climatic Swnmary of the United States" [36-37] (3) The United States Weather Bureau, "Climatography of the United States" [38] (4) The United States Weather Bureau, "Monthly Weather Review" [39] (5) The United States Department of Agricul- ture, "Report of the Chief of Weather Bureau" [40] The characteristics of the precipitation stations are tabulated in Appendix A. b. Runoff records. As a part of Colorado State University hydrology data system, monthly runoff rec- ords have been collected and recorded on magnetic tapes [6,9]. The source of the data is the United States Geological Survey, "Water Supply Papers" [41]. The total number of stations from which data were col- lected is 749. c. Hydrologic data system. There is no relation- ship between the numbering system of runoff stations of the United States Geological Survey and that of precipi- tation stations of the United States Weather Bureau. For fast data processing and particularly for ease of correlation between precipitation and runoff, it is desirable to have identical or almost identical identi- fication numbers for neighboring precipitation and run- off stations for the entire Upper Colorado River Basin. The Colorado State University numbering system was developed for this purpose: --~ (1) Runoff stations are coded with seven digit numbers. Runoff stations within the same drainage have an intermediate number between two limiting num- bers that characterize the downstream and upstream reach of the drainage area [6]. (2) Precipitation stations are coded with eight digit numbers. The first seven digits are identical to the Colorado State University identifica- tion number of the nearest downstream runoff station. However, in some areas there may be several precipita- tion gages close to a single runoff station. The eighth digit in the station number makes it possible to distinguish between the gages in this situation. The precipitation station closest to the associated runoff station is assigned a zero for its eighth digit. The precipitation station next in proximity is assigned one for its eight digit, and so forth. 2. The accuracy of data measurements. It is well known that the observed precipitation does not neces- sarily represent the true amount of water that falls over a station or over the surrounding area [42]. How- ever, the precipitation data that correlate highly with runoff data are still useful indices in this study. 3. Non-homogeneity and inconsistency of records. Non-homogeneity and inconsistency of precipitation data are introduced when there is a change in location, ex- posure, or instrument. Substation History [26-30] and Climatological Data [31-35], both published by the Weather Bureau, show horizontal movement and elevation change. However, the environment and local orography cannot be shown. Most of the drainage area in the Upper Colorado River Basin has been subjected to transmountain diver- sion, transbasin diversion, interbasin diversion, regulation by reservoir, and irrigation diversion that causes a non-homogeneity in the runoff data. The infor- mation about the first four cases is given in the Water Supply Papers [41] and is used for correction of runoff data on the monthly fevel [9]. As to irrigation diver- sion, there is no available record. Furthermore, it is very difficult to estimate seasonable consumptive use and return rate to river. In the high mountain regions, the irrigation allotment is small in amount and is diverted mainly in summer. Correction for irrigation diversion is not done for this reason. 4. Filling missing data. It is necessary to es- tablish a reliable connection between stations having incomplete records and those that are complete. This is done by estimating the missing data from nearby stations with records covering the missing months and having a sufficiently long record which coincides with that of the station with incomplete records. In this study, a simple linear regression method is applied for this purpose. 17