U.S. Department of the Interior Major Components of Flow in the Republican U.S. Geological Survey River During Drought Conditions from Near Hardy, Nebraska, to Concordia, Kansas

Reservoir for use by the KBID is distrib­ streamflow in the Republican River at Background uted by a network of canals that begins Concordia, which is at the downstream just upstream of the gaging station on end of the study area. A companion study Streamflow in the Republican River White Rock Creek at Lovewell. The lands of flow from Concordia to Clay Center, from the Nebraska-Kansas stateline to irrigated by the KBID in the study area Kansas, is being conducted by the Kansas Milford Lake, Kansas (fig. 1), was below are the flat to gently rolling uplands west Geological Survey. normal from March 1988 through June of the Republican River that are drained 1992 because of drought conditions in the by Buffalo and White Rock Creeks and The drought period from March river's drainage basin and probably part of the valley east of the river (fig. 1). 1988 through June 1992 was chosen for because of increased ground- and surface- study to take advantage of the hydrologic water usage. The Republican River is the This fact sheet briefly summarizes and water-use data available for this main source of inflow to Milford Lake; the preliminary results of a study by the period. Monthly major-component and water released from Milford Lake may be U.S. Geological Survey (USGS) to quan­ water-budget estimates were quantified used for industrial, municipal, and agri­ tify those components that have the most using data available from the BOR, the cultural purposes and for maintenance of effect on streamflow in the Republican KBID, the Kansas Department of Agri­ instream uses. Therefore, any loss of flow River during drought conditions from culture, Division of Water Resources in the Republican River decreases the near Hardy. Nebraska, to Concordia, Kan­ (DWR), and the USGS. The monthly amount of water available to replenish sas. A water budget describing the hydro- interval was chosen as a compromise storage in Milford Lake for downstream logic system of flow in this section of the among the varying intervals (daily to uses and may. if the losses are large Republican River was developed using annual) of the available data. enough, cause flow to fall below the mini­ these major components of flow. Monthly mum desirable streamflow (MDS) estimates of the major components of This study was done in cooperation requirement at Concordia, Kansas, set by flow are compared to monthly water-bud­ with the Kansas Water Office and sup­ Kansas law K.S.A. 82a-703a. get estimates, and monthly water-budget ported in part by the Kansas State Water estimates are compared to measured Plan Fund. Following the droughts and floods of the 1930's, the Bureau of Reclamation (BOR) and U.S. Army Corps of Engi­ neers began construction of a series of dams and surface-water irrigation net­ works intended to reduce flooding and to provide water for agriculture. The Kansas Bostwick Irrigation District (KBID) (fig. 1), which was built by BOR and began operation in the study area in 1958, receives most of its water from requested releases from Harlan County Dam in Harlan Nebraska; Harlan County Dam, which County. Guide Rock was completed in 1952, generally does Dam \ \ Belleville Republican not release water unless it is requested by an irrigation district or precipitation is abundant. The releases for the KBID flow down the Republican River and are diverted at Guide Rock, Nebraska, by the Superior-Courtland Diversion Dam (com­ pleted in 1952) into the Courtland Canal, which transports the water to Lovewell Reservoir in Kansas. Lovewell Reser­ voir, which was completed in 1957, gen­ U.S. Geological Survey streamflow- gaging station T I T erally does not release water unless it is 0 5 10 15 KILOMETERS requested by the KBID or precipitation is abundant. Water released from Lovewell Figure 1. Location of Republican River and study area in Kansas. amounts at Belleville, Kansas, were simi­ during the drought of 1952-57 (compare Droughts lar during the three droughts, streamflow figs. 2B and 2C). in the Republican River near Hardy, Climate is of great importance in the Nebraska, generally decreased after Har- study area because most of the area's lan County Dam was completed in 1952. Components of Flow economy is dependent on raising crops and livestock. The climate of the study Although precipitation was greater Components that contribute or area is subhumid (Kansas Water during the drought of 1988-92 than dur­ remove water from a hydrologic system Resources Board, 1961, p. 27), with aver­ ing the drought of 1952-57 (figs. 2B and can be estimated as part of a water bud­ age annual precipitation (1951-80) rang­ 2C, table 1), streamflow was less near get, which can be used to describe the ing from 25 to 29 inches per year from Hardy, Nebraska, and at Concordia, Kan­ system. Components that contribute water west to east (Hedman and Engel, 1989). sas (table 1), during the 1988-92 drought. to the hydrologic system of the study area Recorded annual precipitation at a long- The decrease in streamflow may be due in from outside the area are the Republican term weather station at Belleville, Kansas, part to diversion of water from the Repub­ River itself, the Courtland Canal, and the has varied from 11.79 inches in 1934 to lican River upstream of Hardy for use by movement of ground water through the 49.46 inches in 1993 (fig. 2). This unpre­ the KBID and other irrigators. Stream- adjacent aquifer. Within the study area, dictability and lack of precipitation in flow gains between the gaging stations precipitation is the main source of water some years has led to poor crop yields or near Hardy, Nebraska, and at Concordia, to the hydrologic system. For example, crop failures in the study area. Kansas, during the 1988-92 drought were precipitation may fall directly into water almost double the gains during the 1952- bodies; move by overland flow into tribu­ Droughts occur when precipitation 57 drought (table 1). The larger stream- taries, canals, or the Republican River; or is less than average for several consecu­ flow gains during the 1988-92 drought infiltrate into the ground where it may be tive years (Clement, 1991, p. 288). The probably were caused by canal return used by plants or continue down to the data in figure 2 and table 1 show the two flows from the KBID, which did not water table (fig. 3). Evaporation from regional droughts during \929-4l and begin full operation until 1958. At times water bodies and the land surface and 1952-57 identified by Clement (1991) during the 1952-57 and 1988-92 evaporation and transpiration (evapo- and the drought during 1988-92. Contin­ droughts, the streamflow at Concordia, transpiration) by plants remove water uous precipitation and streamflow data Kansas, was less than the MDS. Stream- from the hydrologic system within the before 1931 are not available for sites in flow at Concordia, Kansas, during Sep­ study area, whereas the Republican River or near the study area. Figure 2A and tember through November 1991 was itself and ground-water movement table 1 show that, although precipitation much less than the MDS than at any time through the adjacent aquifer remove

(A) OCTOBER 1931-SEPTEMBER 1995 i i 1 i . . . 1 i i . . , . . . . 1 . . , , 1 i , , i , , . . | , , , , | , , . | . . . . | = 50 1 ' L co g 40 . At Belleville, Kansas 1 ^ *?n EXPLANATION 2 ^20 PRECIPPERYEAR ^. 10 Precipitation - Average annual precipitation 0 ,, 1 ,,,,!, , , 1 , , , 1 , , , , 1 , , , , 1 , , , i 1 , , , , 1 , , , , 1 i i , , 1 i i i i 1 (1932-51 and 1952-95) 2,000,000 _' ' i ' ' i i I i i ,,,,,,,,, i i i |_ STREAMFLOW, 1,000,000 ^^Harlan, County Dam completed INACRE-FEET 500,000 Average annual gaged 200,000 '_ Near Hardy, Nebraska streamflow (1932-51 and 100,000 x- Kansas Bostwick Irrigation 50,000 / District completed Gaged streamflow 1952-95) 20,000 mnnn " i i 1 , , 1 i , , i , , , i i , i i , r 1932 1935 1940 1945 1950 1955 1960 1965 1970 1975 1980 1985 1990 1995

(B) OCTOBER 1951-SEPTEMBER 1957 (C) OCTOBER 1987-SEPTEMBER 1992

gS 10 At Concordia, Kansas £co ^ 8 2 LLJ t; /Precipitation /; EXPLANATION a. o^ o? 6 s Precipitation ^3^^ ^ -Average monthly precipitation £^ 2 n (1952-95) 0 100,000 1 ' I 1 """ ' ' ' ' 1 '">'»""| iiiiinnii |i, ,11 in, i M in , i Sft! 50,000 3 £ £ 20,000 - . Minimum desirable streamflow S= ih o 10,000 < £ 2 5,000 cc < £ 2,000 fe ^ °- 1,000 _ Gaged streamflow Gaged streamflow Fnn r.nl.iiiniT 1951 1952 1953 1954 1955 1956 1957 1987 1988 1989 1990 1991 1992

Figure 2. Precipitation and gaged streamflow in the study area: A, Average annual, October 1931-September 1995; B, Average monthly, October 1952-September 1956; C, Average monthly, October 1987-September 1992. (Sources: precipitation data from National Climatic Data Center, Asheville, North Carolina, 1996; streamflow data from U.S. Geological Survey, Lawrence, Kansas, 1996) Table 1. Comparison of average annual precipitation and streamflow during three of flow used to describe flow in the drought and three reference periods in the study area Republican River in the study area are apstreomflow near Hardy, Nebraska, plus [Precipitation is average annual precipitation in inches per year; streamflow is average annual streamflow in acre-feet per year; --, no data; purple shading indicates drought periods. Precipitation data from National tributary flow (including municipal and Climatic Data Center, Asheville, North Carolina, 1996. Streamflow data from U.S. Geological Survey, industrial discharges), plus Lawrence, Kansas, 1996] seepage plus canal return jft&*. minus surface-water diversions. Environmental Time periods factors and locations 1932-1941 1942-1951 1952-1957 1958-1987 1988-1992 1993-1995 Upstream flow data were collected and measured by the USGS at a stream- Precipitation flow gaging station near Hardy, Nebraska. Belleville, Kansas 24.97 32.74 23.12 31.87 24.15 36.97 These data provided a measurable inflow Concordia, 18.89 29.42 24.25 32.76 for purposes of the water budget. Because Kansas water in the Republican River takes about Streamflow 1 day to travel from the gaging station Near Hardy, 517,784 763,296 238.863 273,253 100,662 407,018 near Hardy, Nebraska, to the gaging sta­ Nebraska tion at Concordia, Kansas, the upstream Concordia. 285,488 450,245 184,036 790,029 flow used was for the day previous to that Kansas measured at Concordia. water from the system to outside the study small and was ignored. Other minor com­ Tributary flow includes streamflow area. Water diverted within the study area ponents were evaporation from water data collected and measured by the USGS by humans may be consumed and bodies and the land surface, evapotranspi- at gaging stations on White Rock Creek at removed from the system, but some of the ration from plants, and infiltration to the Lovewell. Kansas, and on Buffalo Creek diverted water may be returned to the sys­ water table from all sources other than near Jamestown, Kansas; estimates of tem by infiltration to the water table or seepage across the streambed of the ungaged flow, which were made on the through discharges into tributaries, Republican River (fig. 3). basis of statistical methods; and discharge canals, or the Republican River. of water from municipal and industrial Water-Budget Estimates sources. Monthly discharge from munici­ The water budget used in this study pal and industrial sources was estimated describes flow in the Republican River Each of the components of flow used from annual discharge values and only and not the hydrologic system of the in the water budget were quantified on a monthly diversion values reported by the entire study area. Therefore, not all com­ monthly basis for the drought period dur­ municipality or industry to the DWR. ponents that contributed water to or ing March 1988 through June 1992. removed water from the study area (fig. 3) These quantified components then were Streambed seepagt is defined in this were considered part of the water budget combined to make monthly estimates of study as the increase or decrease of flow describing the flow in the Republican flow in the Republican River at Concor­ in the Republican River from seepage River. Water that enters or leaves the dia, Kansas, during this drought period. between the aquifer and the river. Stream- Republican River in the study area The water budget and major components bed seepage was estimated using the com- through tributaries or canals, through exchange of water between the aquifer and the river by streambed seepage, or through use by humans were considered to be major components of flow to the Republican River and were used in the water budget. Other components of flow Evapotranspiratron Overland were considered minor and were not used runoff Surface-water in the water budget because they are diversion included within the major components or are small in comparison. For example, most of the precipitation that falls within the study area becomes part of other components (for example, flow in tribu­ taries) before it reaches the Republican River and, therefore, is accounted for in the contributions made by the major com­ NOT TO SCALE ponents; the remaining precipitation that falls directly on the Republican River is Figure 3. Components of flow in the study area. puter program BFI4 (Wahl and Wahl, Table 2. Monthly water-budget estimates of flow in the Republican River from near 1995) that computes daily streambed Hardy, Nebraska, to Concordia, Kansas, compared to measured downstream flow and seepage from average daily streamflow. minimum desirable streamflow at Concordia, Kansas, during October 1990 through The average daily streamflow data col­ May 1992 lected and measured by the USGS at gag­ [All values are in acre-feet. Minimum desirable streamflow from Kansas law K.S.A. 82a-703a Blue shading ing stations at Concordia, Kansas, and indicates critical irrigation period; purple shading indicates peak of drought period] near Hardy, Nebraska, were used as input Water- Measured Minimum to the program. The BFI4 program uses budget down­ desirable Sur­ estimate stream stream- streamflow hydrograph-separation tech­ face- of flow at flow at Up­ . Tribu- . Stream- Canal water _ down­ Con­ Con­ niques to estimate the part of the flow in stream H tary + bed + return diver­ stream cordia, cordia, Year Month flow flow seepage flow sions flow Kansas Kansas an unregulated stream that is from stream- 1990 October 2,150 203 1,715 0 0 4,068 3,955 5,227 bed seepage. The Republican River is November 2,138 286 1,197 0 0 3,621 3,495 4,760 regulated by Harlan County Reservoir. December 2,172 293 1,197 0 0 3.662 3.420 6,149 Under low-flow conditions, releases from 1991 January 3.800 549 1,573 0 0 5,923 5,260 6,149 Harlan County Reservoir typically are February 8,255 1,985 637 0 0 10,877 11,151 6,942 fully diverted into the Courtland Canal at March 4,274 1,665 2,265 0 0 8,204 6,538 9,223 Guide Rock, Nebraska. Therefore, for the April 2,313 3,320 1,567 0 0 7,200 4,514 8,926 purposes of this study, the Republican May 5,361 6,083 1.627 0 0 13.072 6.849 9.223 River can be considered to begin immedi­ June 14.236 6.458 2.687 1.096 1,360 23,117 26,918 8,926 ately downstream of the Guide Rock July 3,320 210 2,286 2,655 3,756 4,715 5,435 9,223 diversion (Thomas Stiles, Kansas Water August 3,600 156 1,522 1,264 1,360 5,183 5,282 9,223 Office, oral commun., 1996). Under these September 920 151 26 0 0 1,097 1,864 4,760 conditions, the Republican River in the October 1,033 129 -199 0 0 964 895 5,227 study area may be viewed as unregulated November 1.333 88 494 0 0 1,914 2,021 4,760 and suitable for application of the BFI4 December 1,595 134 1,169 0 0 2,897 3,013 6,149 program. The amount of streambed seep­ 1992 January 2,075 231 2,120 0 0 4,427 4,267 6,149 age to and from the Republican River was February 1,551 185 1,923 0 0 3,659 3,550 6,942 estimated as the BF14 results for the gag­ March 4,417 580 1,933 0 0 6,930 5,794 9.223 ing station near Hardy, Nebraska, sub­ April 3.739 753 2.208 0 0 6,700 5,798 8,926 tracted from the BF14 results for the May 1,835 271 1,296 0 0 3,402 3,148 9,223 gaging station at Concordia, Kansas, after Total 70,117 23,730 29,243 5,015 6,476 121,632 113,167 145,330 compensating for a 1-day traveltime. Monthly average 3,506 1,186 1,462 251 324 6,082 5,658 7,266 C&nel return flov data were col­ lected and measured by the KB ID and Jamestown; these diversions already are ber 1990 through May 1992). The second BOR at discharge gates at the end of accounted for in the tributary or canal- period was at the peak of the drought canals in the KBID. Canal return flow return-flow components. Only diversions when downstream flow was extremely from the KBID to the Republican River from tributaries and canals in the ungaged low (September and October 1991). The generally occurs only during the months part of the study area or from the Republi­ third period was the critical irrigation that irrigation is allowed in the district can River in ihe study area were consid­ period (July and August 1991) when (June through September). Although irri­ ered to be part of this component. All upstream and downstream flow was low gation does not occur in all 4 months dur­ water from these surface-water diversions and irrigation demand was high. Annual ing all years, canal return flow has was used for irrigation. The annual sur­ precipitation at Concordia, Kansas, was occurred in July and August every year face-water irrigation-use data for these 70, 37, and 49 percent of average (1952- since 1958. diversions were divided into monthly data 95), respectively, during these three peri­ in the same ratios as the monthly irriga­ ods. Table 2 shows the monthly estimates Surjace-water divenions? the tion data available from the KBID. of the major components of flow and the amount of surface water removed water-budget estimates for October 1990 (diverted) by humans, were estimated through May 1992 (which includes peak from annual water-use data reported to Comparison of Water-Budget of the drought and critical irrigation peri­ DWR by the users. The only uses for Estimates ods). Measured downstream flow and which surface water was diverted in the MDS are included in table 2 for compari­ study area were irrigation and recreation. Three important periods within the son purposes. Each month's water-budget Many of the diversions (including the March 1988 through June 1992 drought estimate of downstream flow can be com­ large diversions to and from Lovewell were identified. The first period was when pared with the monthly measured down­ Reservoir for the KBID) were upstream measured streamflow at the gaging station stream flow (table 2). Table 3 shows from the gaging stations on White Rock at Concordia, Kansas (downstream flow), comparison of estimates of each of the Creek at Lovewell or on Buffalo Creek at commonly was less than the MDS (Octo­ major components of flow and of mea- Table 3. Comparison of major components of flow and downstream flow to water- cent aquifer. During the critical irrigation budget estimates of downstream flow during important periods in 1988-92 drought period, Streambed seepage contributed about 38 percent of the water-budget esti­ [MDS, minimum desirable streamflow at Concordia, Kansas (Kansas law K.S.A. 82a-703a). Components of mates. Although Streambed seepage can flow may nol add up to 100 percent due to errors in estimation] contribute substantially to the water-bud­ Component as a percentage of water-budget estimate of downstream flow get estimates during most of a drought, it Peak of Critical cannot be considered a reliable source of Period with downstream flow droughtarougni period irrigation periodpenuc water during periods of very low flow. generally less than MDS (September-" ' - (July-August' 9 ' (October 1990-May 1992) October 1991) 1991) Major Monthly Monthly Canal return flow and surface- component Monthly range average average Monthly average water diversions occur only during the Upstream flow 32 to 107 58 95 70 irrigation season (commonly the months Tributary flow 3 to 46 20 14 4 of June through September); therefore, -8 Streambed seepage -21 to 52 24 38 they can seem to be insignificant compo­ Canal return flow 0 to 56 4 0 40 nents of the water-budget estimates if Surface-water diversions OtoSO 5 0 52 averaged over a period longer than the Measured downstream flow 52 to 170 93 134 108 irrigation season. For example, canal Diversions are removals of water from the system and are substracted when combining the major com­ return flow contributed about 4 percent to, ponents of flow to make the water-budget estimates. and surface-water diversions removed about 5 percent from, the water-budget sured downstream flow to the water-bud­ water budget but averaged about 20 per­ estimates if averaged over the period of get estimates of downstream flow during cent due to large contributions in some October 1990 through May 1992; nothing each of the three periods identified within months (tables 2 and 3). Tributary flow was contributed to or removed from the the March 1988 through June 1992 averaged a little less, about 14 percent of water-budget estimates by these compo­ drought. the water-budget estimates, during the nents during the peak of the drought (Sep­ peak of the drought. However, during the tember and October 1991). However, Upstream flow is generally the larg­ critical irrigation period, the contribution during the critical irrigation period (July est contributor to the water-budget esti­ of tributary flow was very small, averag­ and August 1991), canal return flow con­ mates. Even during the period from ing only about 4 percent of the water-bud­ tributed about 40 percent to and surface- October 1990 through May 1992 when get estimates. Although tributary flow can water diversions removed about 52 per­ upstream flow was only about 16 percent contribute substantial amounts of water to cent from the water-budget estimates. The of average (1952-95) and downstream the water-budget estimates, it cannot be combination of these two components can flow commonly was less than the MDS, considered a reliable source of water dur­ be considered to show the effect of human upstream flow typically contributed more ing the critical irrigation months of July use (irrigation) on the water-budget esti­ than one-half and averaged about and August or during periods when pre­ mates in the ungaged part of the study 58 percent of the water-budget estimates cipitation is consistently less than about 2 area. In 10 out of the 15 months that irri­ (tables 2 and 3). During the peak of the inches per month. gation is estimated to have occurred dur­ drought (September and October 1991), ing the March 1988 through June 1992 upstream flow was only about 4 percent Streambed seepage generally is pos­ drought, the combination of the canal of average (September and October itive, indicating that the Republican River return flow and surface-water diversions 1952-95), but its contribution to the is a gaining stream that is, the ground resulted in a small net contribution of water-budget estimates increased to about water flows into the river from the adja­ about 1 percent to the water-budget esti­ 95 percent. During the critical irrigation cent aquifer because the water table is mates of downstream flow. However, dur­ period (July and August 1991), upstream higher than the water level in the river ing the 1991 irrigation season, the flow was about 13 percent of average (fig. 3; tables 2 and 3). A negative stream- members of the KB ID were restricted in (July and August 1952-95) and contrib­ bed-seepage value may indicate that this their use of water, and surface-water uted about 70 percent of the water-budget situation is reversed and that the water diversions by other users were greater estimates. table in the adjacent aquifer is lower than than in previous years. The combination the water level in the river causing the of the canal return flow and surface-water Tributary flow can be a major con­ river to lose water to the aquifer (for diversions during July and August 1991 tributor to the water budget, especially example, October 1991, table 2). Stream- resulted in a net withdrawal of about 12 when evapotranspiration is low (winter bed seepage averaged about 24 percent of percent from the water-budget estimates. and early spring) or when precipitation is the water-budget estimates during Octo­ consistently about 2 or more inches per ber 1990 through May 1992. However, Downstream flow data were col­ month (for example, April through June during the peak of the drought when lected and measured by the USGS at a 1991). However, during October 1990 streamflow was very low in the Republi­ streamflow-gaging station on the Republi­ through May 1992, tributary flow com­ can River, Streambed seepage averaged can River at Concordia, Kansas. These monly was less than 10 percent of the -8 percent, indicating a loss to the adja­ data provided a measurable outflow for comparison with the water-budget esti­ the water-budget estimates during the be managed to minimize the effect of mates. Downstream flow was about periods when downstream flow com­ drought conditions may allow streamflow 16 percent of average (1952-95) during monly was less than the MDS and when in the Republican River at Concordia. the drought period, and commonly was the drought peaked. However, it was not a Kansas, to remain above the MDS and less than the MDS. Downstream flow significant component during the critical ensure an adequate supply of water to ful­ decreased to about 4 percent of average irrigation period and, therefore, is not fill the needs of downstream users. (September and October 1952-95) during considered a reliable source of water to the Republican River during drought peri­ the peak of the drought period (table 2). Cristi V. Hansen Tables 2 and 3 show that although there ods. Streambed seepage was an important was substantial disagreement between the component of the water-budget estimates water-budget estimates and downstream during the period when downstream flow flow from month to month during October commonly was less than the MDS and Selected References 1990 through May 1992, monthly water- during the critical irrigation period. How­ ever, the value of this component can budget estimates varied on average less Clement, R.W., 1991, Kansas floods and become negative, as it did during the peak than 10 percent from measured down­ droughts, in Paulson, R.W., Chase, of the drought period, resulting in remov­ stream flow during both the period when E.B., Roberts, R.S., and Moody, als from rather than contributions to the downstream flow commonly was less D.W., compilers, National water water-budget estimates. This indicates than the MDS and the critical irrigation summary 1988-89 Hydrologic that streambed seepage is not always an period. However, during the peak of the events and floods and droughts: U.S. available or reliable source of water to the drought period, measured downstream Geological Survey Water-Supply Republican River because at times the Paper 2375, p. 287-294. flow exceeded water-budget estimates of river loses water to the adjacent aquifer. downstream flow by about one-third: this Canal return flow and surface-water Hedman, E.R., and Engel, G.B., 1989, may be due to errors in estimating the diversions were small if averaged over the Flow characteristics for selected tributary flow. period when downstream flow was less streams in the Great Plains subregion than the MDS and were zero during the of the Central Midwest Regional period at the peak of the drought; how­ Aquifer System and selected adja­ Conclusions ever, they accounted for large contribu­ cent areas Kansas and Nebraska, tions to and removal from the water- and parts of Colorado, Iowa. Mis­ souri, New Mexico, Oklahoma, budget estimates during the critical irriga­ During the drought of March 1988 South Dakota, Texas and Wyoming: tion period. Although the combination of through June 1992, there were three U.S. Geological Survey Hydrologic important periods the period when flow canal return flow and surface-water diver­ Investigations Atlas HA-708, in the Republican River at Concordia, sions commonly results in a net contribu­ 3 sheets, scale 1: 1,000,000. Kansas, commonly was less than the tion to the water-budget estimates, when MDS (October 1990 through May 1992); water usage by the KBID is restricted, as Kansas Water Resources Board, 1961, the period at the peak of the drought when it was during the critical irrigation period, Part A Preliminary appraisal of flow in the Republican River was very their combination may become negative Kansas water problems section 9, low (September and October 1991); and and result in a net withdrawal. Therefore, Lower Republican unit: Topeka, the critical irrigation period (July and canal return flows and surface-water Kansas, State Water Plan Studies, August 1991). diversions are each considered important 99 p. components of the flow in the Republican Wahl, K.L., and Wahl, T.L., 1995, Deter­ During all three periods, the River during irrigation seasons, but their combination may change from a net con­ mining the flow of Comal Springs at upstream flow was the most important New Braunfels, Texas, in Proceed­ tribution to a net withdrawal of water and most reliable component of the ings of Texas Water '95, A Compo­ from the river if water usage by the KBID monthly water-budget estimates. nent Conference of the First Upstream flow is considered the most is restricted. International Conference on Water important and reliable source of water to Resources Engineering: American the Republican River, with its importance Consideration of upstream flow, Society of Civil Engineers, August increasing as the drought worsens. Tribu­ canal return flow, and surface-water 16-17, 1995, San Antonio. Texas, tary flow was an important component of diversions and how these components can p. 77-86.

For more information, please contact: District Chief U.S. Geological Survey 4821 Quail Crest Place Lawrence, Kansas 66049 (913)842-9909 email: [email protected]

March 1997 Fact Sheet FS-234-96