Hydrologic Droughts in — Are They Becoming Worse? By James E. Putnam, Charles A. Perry, and David M. Wolock

Statewide median runoff for Kansas for water years 1920–2006 Introduction (fig. 2) indicates large departures from median runoff during five historic hydrologic droughts in Kansas that occurred during Multi-year droughts have been a recurrent feature of the 1929–41, 1952–57, 1962–72, 1974–82, and 1988–92 (Paulson climate and hydrology of Kansas since at least the 1930s. and others, 1991). The severity of streamflow conditions during Streamflow records collected by the U.S. Geological Survey water years 2000–06 is apparent when (USGS) indicate that water years 2000 to 2006 (October 1, comparing the departures from median 1999, through September 30, 2006) represent the sixth hydro- annual runoff for that period with those logic drought during the past eight decades, and that corre- during the historic droughts. Annual sponding streamflow levels in some parts of Kansas were lower runoff in Kansas was below the median than those during historic droughts of the all years during 2000–06 except water 1930s and 1950s, even though the year 2004, which was near the median. precipitation deficit was not as Water year 2006, the driest year during severe. Record-low stream- the 2000–06 hydrologic drought, had flows in water year 2006 the second lowest annual runoff during were recorded at USGS 1920–2006; the lowest annual runoff streamgages on the for Kansas occurred in water year Republican, Smoky 1956. Hill, Solomon, Saline, The streamflow effects of drought upper Kansas, middle during water years 2000–06 were Arkansas, and Little focused in northern Kansas on the Arkansas , as Republican, Saline, Solomon, and well as many tributary Smoky Hill Rivers (fig. 1), whereas the sites, and one tributary 1950s drought effect was focused more site of the Neosho in southeastern Kansas on the Verdi- (fig. 1, table 1). gris, Fall, and Neosho Rivers (Paulson Low streamflows dur- and others, 1991). Average annual ing the hydrologic drought also flows at four long-term streamgages resulted in record low levels at (fig. 3) indicate that average annual three Federal reservoirs in Kansas streamflow during 2000–06 at each (fig. 1, table 2). An unprecedented number of administrative of the streamgages was less than the decisions were made by the Division of Water Resources, average annual streamflow during other Kansas Department of Agriculture to curtail water diversions historic droughts in Kansas. Further- from rivers to maintain minimum desirable streamflows, and more, the average annual flow for all low flows on the lower Republican River in Kansas created streamgages was less than 25 percent concerns that Colorado and Nebraska were not complying with of long-term median annual flow dur- the terms of the 1943 Republican River Compact. ing 2000–06, indicating the severity of the drought. For example, annual average streamflow for the Republican Statewide Runoff and Streamflow River at Clay Center (site 5, fig. 1, fig. 3A, table 1), in continuous opera- Conditions tion since 1919, was less than the long- term median streamflow each water Daily streamflow data for water years 1900–2007 are used year since 2000 and was the lowest on record in water year to estimate average annual runoff (streamflow per unit area) for 2006. Moreover, five of the 10 lowest annual average flows for all streamgages in Kansas (U.S. Geological Survey, 2006a). the Clay Center gage occurred during 2000–06 (U.S. Geologi- The median runoff for each water year is then determined. cal Survey, 2006b).

U.S. Department of the Interior Fact Sheet 2008–3034 U.S. Geological Survey Printed on recycled paper April 2008 102° 101° 96° 40° 100° 99° 98° 97°

95° Big Republican Cr. White Little Creek Creek Delaware Fork River Creek 1 Rock Sappa 3 South Beaver 2 Dog Blue Republican Little Beaver North Fork 19 MISSOURI Fork 14 Stranger Fork Middle Prairie Fork 4 South River South Perry River RIVER 15 Lake

Saline 5 Creek River Milford 6 RIVER 20 River KANSAS 39° Saline 12 River Lake Smoky 11 16 R. River 18 Wakarusa Ladder 7 Smoky Kanopolis River 17 Hill Lake 10 13 Marais Whitewoman Creek A 8 9 Cygnes Hill des River Creek Neosho B Creek Little Cottonwood Walnut A River

21 ArkansasB Pawnee River 22 ARKANSAS Creek 38° ARKANSAS RIVER North Fork Fall River A Cr. Verdigris Crooked South Fork Ninnescah Walnut River Bear Fork River B RattlesnakeMedicine River Lodge Chikaskia River River River North RIVER 23

Creek River River Cimarron 37° Cimarron River Base from U.S. Geological Survey digital data, 1:2,000,000, 1994 0 20 40 60 MILES Albers Equal-Area Conic projection Standard parallels 29°30' and 45°30', central meridian -96°00' EXPLANATION 0 20 40 60 KILOMETERS

Drought condition by hydrologic unit 18 Streamgage with record-low annual Drainage basin average streamflow—Map index A Basin Extreme Below normal number shown in table 1 Severe Normal B Lower Basin 20 Reservoir with record-low elevation— Moderate Map index number shown in table 2 Basin boundary

Figure 1. Drought conditions by hydrologic unit in Kansas, comparison of water year 2006 streamflows with all other annual streamflows, water years 1930–2006, and location of streamgage with record-low annual streamflow in water year 2006, and reservoirs with record-low elevations, water years 2000–06 (hydrologic units from Seaber and others, 1987; data for drought conditions from WaterWatch, 2006 (U.S. Geological Survey, 2007a).

Period of hydrologic drought 1929–41 1952–57 1962–72 1974–82 1988–92 2000–06 Current and Historic 6 Precipitation ,

f 5 f

o Annual

n departure u Meteorological conditions r

l

a 4 during the 2000–06 drought were u n

n different than those during previ- a r

a 3 e ous severe droughts in the 1930s y

5-year r

e moving and 1950s. Statewide average median p

average s 2 m annual precipitation for Kansas e o h r f c during water years 2000–06 n e i r

u n 1 i t (fig. 4) was greater than that dur- r Median a

p ing all previous droughts except e d

l 0 for the 1974–82 drought (National a u n

n Oceanic and Atmospheric Admin- A -1 istration, 2007a). Furthermore, statewide average annual precipi- -2 0 5 0 5 0 5 0 5 0 5 0 5 0 5 0 5 0 5 0 tation for water years 2000–06 2 2 3 3 4 4 5 5 6 6 7 7 8 8 9 9 0 0 1 9 9 9 9 9 9 9 9 9 9 9 9 9 9 9 9 0 0 0 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 2 2 2 was less than 1 inch below nor- Water year mal, whereas the average annual precipitation during one of the Figure 2. Statewide runoff showing annual departure of runoff from the statewide median of 1.63 worst Kansas droughts, 1952–57, inches per year and a 5-year moving average line for water years 1920–2006. Wetter years plot above was nearly 6 inches below normal. the median line and drier years plot below the median line, (data from U.S. Geological Survey, 2006a). The average statewide Palmer Table 1. U.S. Geological Survey streamgages with record-low average annual streamflows during water year 2006. [Data from U.S. Geological Survey, 2006b; mi2, square miles; ft3/s, cubic feet per second] Map index Long-term median Water year 2006 Drainage Years of number Station name annual average annual average area (mi2) record (fig. 1) streamflow (ft3/s) streamflow (ft3/s) 1 Prairie Dog Creek near Woodruff, KS 1,007 0.11 0.03 61 2 White Rock Creek near Burr Oak, KS 227 22.5 2.52 49 3 Republican River near Hardy, NE 22,401 380 19.2 74 4 Republican River at Concordia, KS 23,560 492 44.9 60 5 Republican River at Clay Center, KS 24,542 827 82.7 88 7 Big Creek near Hays, KS 549 18.4 .71 60 8 at Ellsworth, KS 7,580 151 24.2 94 10 Smoky Hill River near Mentor, KS 8,358 274 26.8 65 11 Saline River near Russell, KS 1,502 47.7 3.59 55 12 Saline River at Tescott, KS 2,820 120 14.1 86 13 Smoky Hill River at New Cambria, KS 11,730 398 60.3 44 14 North Fork Solomon River at Portis, KS 2,315 77.8 11.8 54 15 at Osborne, KS 2,012 42.0 4.10 61 16 Solomon River at Niles, KS 6,770 328 72.0 95 17 Smoky Hill River at Enterprise, KS 19,260 1,139 191 72 18 at Fort Riley, KS 44,870 2,019 418 42 19 Little Blue River near Barnes, KS 3,351 629 213 48 21 Cow Creek near Lyons, KS 728 53.4 7.42 58 22 Little Arkansas River at Alta Mills, KS 736 168 11.8 33 23 Lightning Creek near McCune, KS 197 131 10.7 55

Drought Severity Index (PDSI), one of the indices used to Runoff-Precipitation Relation classify and assess long-term meteorological droughts (Hayes, 2002), for water years 2000–06 was -0.03 (normal), whereas The Statewide relation between runoff and precipitation the average PDSI for the 1952–57 drought was -3.22, indicating (fig. 5) does not indicate that there has been a significant change a severe drought (National Oceanic and Atmospheric Admin- in runoff between past droughts and the 2000–06 drought when istration, 2007b). Statewide average precipitation during the comparing the lowest five consecutive years of runoff in each five consecutive driest years of the 2000–06 drought also was drought. This statewide analysis, however, includes many areas greater than the statewide average precipitation during the five of Kansas that were not even in drought conditions during driest years of the 1930s and 1950s droughts (fig. 5); however, much of 2000–06. The relation indicates that, generally, runoff the annual statewide median runoff during the 2000–2006 increases as precipitation increases, an intuitive conclusion. drought was near the low runoff levels that occurred in the However, if the same analysis is completed using median annual 1930s and the 1950s. Precipitation in the area of Kansas where runoff for the four long-term sites in Kansas (fig. 3) that repre- most of the record-low streamflows occurred during water year sent the worst drought-affected areas during the 2000s drought, 2006 was generally about 75–90 percent of normal, not an indi- the results appear different (fig. 6) because the regional affects cation of a significant lack of precipitation (National Oceanic of the 2000s drought are not smoothed by statewide averaging and Atmospheric Administration, 2007a).

Table 2. Location of Federal reservoirs in Kansas with record low elevations during water years 2000–06. [Data from Putnam and Schneider, 2003 and U.S. Geological Survey, 2006c; mi2, square miles; ft, feet] Map index Drainage Top of conserva- Percent number Station name area tion pool Lowest elevation and date conservation (fig. 1) (mi2) elevation (ft) pool Elevation (ft) Water Year 6 Milford Lake near Junction City 24,880 1,144.4 1,136.90 2003 68 9 Kanopolis Lake near Kanopolis 7,857 1,463 1,456.22 2006 55 20 Perry Lake near Perry 1,117 891.5 884.90 2003 64 A. Republican River at Clay Center (site 5, fig. 1, table 1, 88 years of record) 28 1,000 3 26.88 Long-term median annual flow 827 ft /s 27 Normal precipitation 26.68 26.43 800 1920–2006 26 (27.55 inches) 25.90 600 25 24.54 400 24 200 23 0 B. Saline River at Tescott (site 12, fig. 1, table 1, 86 years of record) 22 21.61 150 Long-term median annual flow 120 ft3/s 21 Average annual precipitation, in inches Average /s) 3 100 20 1929–41 1952–57 1962–72 1974–82 1988–92 2000–06 Periods of hydrologic drought (water years) 50 Figure 4. Statewide average annual precipitation during water 0 years 2000–06 compared with average annual precipitation C. Solomon River at Niles (site 16, fig. 1, table 1, 95 years of record) during five historic droughts in Kansas (precipitation data from 400 National Oceanic and Atmospheric Administration, 2007a). Long-term median annual flow 328 ft3/s 300 five driest consecutive years for each drought (same water years 200 as those used in previous analysis) was much lower during the 2000s drought than during the 1930s and 1950s drought, even 100 though average precipitation was greater (fig. 6). This analysis indicates a change in the runoff-precipitation relation between

Average annual streamflow, in cubic feet per second (ft annual streamflow, Average 0 earlier droughts and the current drought. D. Smoky Hill River at Enterprise (site 17, fig. 1, table 1, 72 years of record) The runoff ratio, the amount of precipitation that becomes 1,500 streamflow, can be computed annually for each of the sites Long-term median annual flow 1,139 ft3/s listed above by dividing the annual runoff by annual precipita- tion and multiplying by 100. A 5-year moving average was 1,000 used to smooth the computed annual runoff ratios. For each site, the mid year of the five consecutive driest years during 500 each drought was determined for each drought. For each site, the runoff ratio has decreased for each drought, and the more 0 recent droughts have a lower runoff ratio, as shown in figure 1929–41 1952–57 1962–72 1974–82 1988–92 2000–06 7. The runoff ratio for the 2000s drought was the lowest of all Period of hydrologic drought (water years) droughts.

Figure 3. Average annual streamflow at four selected 1.4 streamgages in Kansas during five historic droughts and for water 1963−67 1977−81 years 2000–06 (data from U.S. Geological Survey, 2006b). 1.2

1.0 as in figure 5. Streamflow records have been collected since 2002−06 the 1930s at Republican River at Clay Center (site 5, fig. 0.8 1, table 1), Saline River at Tescott (site 12, fig. 1, table 1), 1988−92 0.6 1936−40 Solomon River at Niles (site 16, fig. 1, table 1), and Smoky 1952−56 Hill River at Enterprise (site 17, fig. 1, table 1). Moreover, these sites are located at the downstream most location in the 0.4 primary river basins that experienced record-low streamflows 0.2 in water year 2006. Precipitation data for this analysis is from Average median runoff, in inches per year Average the PRISM group (PRISM, 2007), which provides precipitation 0 data that are distributed over a delineated river basin. National 19 20 21 22 23 24 25 26 27 28 Oceanic and Atmospheric Administration divisional precipi- Average annual precipitation, in inches per year tation data (used in analysis in figure 5) is collected at nine nearly equal areas in Kansas, and does not provide data for a Figure 5. Statewide relation between median annual runoff and specific basin. In fact, these precipitation data may include data average annual precipitation for 5 consecutive driest years during for parts of several river basins; therefore, it was not used. six hydrologic droughts in Kansas (precipitation data from National The relation between annual precipitation for the river Oceanic and Atmospheric Administration, 2007a; runoff data from basin, and averaged annual runoff for the four sites during the U.S. Geological Survey, 2006a). 0.6 Many other factors in addition to those discussed in the 1952−56 preceding paragraphs, or a combination of factors, could be 1963−67 1977−81 0.5 affecting the historical relation between runoff and precipitation, 1936−40 the change in runoff ratio, and the low streamflows documented 0.4 1988−92 during the 2000s hydrologic drought. A better understanding of the runoff ratio change and the contribution of each of the fac- tors needs more study and is beyond the scope of this paper, but 0.3 is critical to improved drought management. 2002−06 0.2 A. Republican River at Clay Center (site 5, fig. 1, table 1) 0.1 4.0

Median annual runoff, in inches per year 3.5 0 3.0 16 17 18 19 20 21 22 23 1938 2.5 Average annual precipitation, in inches per year 1965 2.0 1979 Figure 6. Relation between annual runoff and annual precipitation 1.5 for five consecutive driest years during six hydrologic droughts at 1954 1.0 Republican River at Clay Center, Saline River at Tescott, Solomon 1990 River at Niles, and Smoky Hill River at Enterprise (precipitation data 0.5 2004 from PRISM Group, 2007; runoff data from U.S. Geological Survey, 0 B. Saline River at Tescott (site 12, fig. 1, table 1) 2006b). 4.0 3.5 1954 The decreases in runoff-precipitation ratio for the four 3.0 1938 sites in figure 7 indicate that less precipitation was contributed 2.5 to streamflow during the 2000s drought then in the previous 2.0 1965 1990 droughts, and that this change has occurred progressively with 1979 time. What may contribute to this change? One of the most 1.5 1.0 significant water- and land-management changes in Kansas 2004 has been the construction of a large number of small lakes 0.5 used for water supply and flood control. The number of small 0 lakes in Kansas has increased from 12 in 1920 to 5,240 in 1995 C. Solomon River at Niles (site 16, fig. 1, table 1) (USACE, 2007). For example, nearly 78 percent of the basin 4.0 1954 upstream from Solomon River at Niles (site 16, fig. 1, fig. 7C, 3.5 1938 1965 table 1) is affected by impoundments (Putnam, 2003). These Runoff ratio, in percent 3.0 1979 impoundments trap runoff that may be slowly released, or the 2.5 1990 water is lost by infiltration or evaporation, and never reaches a 2.0 stream channel. Other conservation and farming practices, such 1.5 as contour farming and terraces, also may contribute in part to 2004 1.0 the change in runoff ratio because these land-management prac- tices trap runoff, allowing it to infiltrate slowly into the ground 0.5 instead of flowing into streams. Other changes in land manage- 0 D. Smoky Hill River at Enterprise (site 17, fig. 1, table 1) ment, such as conversion of crop land to grassland (Conserva- 4.0 tion Reserve Program), also may affect the runoff-precipitation 3.5 1954 1979 ratio. 3.0 1965 Demand for water has increased in Kansas and is expected 1938 to continue to increase. Changes in water use also may be a 2.5 factor in the runoff ratio decrease. Increases in water demand 2.0 become a management issue during periods of drought (Nace 1.5 1990 2004 and Pluhowski, 1965). Water use in Kansas increased 200 1.0 percent between 1955 and 1980 (Joan Kenny, USGS, written 0.5 comm., 2007), largely because of the development of ground- 0 water wells for irrigation in western Kansas. Water use in the 1930 1940 1950 1960 1970 1980 1990 2000 2010 State decreased from 6,600 million gallons per day in 1980 to 3,670 million gallons per day in 2005 (Joan Kenny, USGS, Figure 7. Runoff ratio at (A) Republican River at Clay Center, (B) written comm., 2007). In 2006, communities that receive their Saline River at Tescott, (C) Solomon River at Niles, and (D) Smoky municipal water supply from alluvial wells connected to the Hill River at Enterprise for the mid-year of five consecutive driest Smoky Hill River experienced water shortages because of low years at six hydrologic droughts in Kansas (precipitation data from streamflows. PRISM group, 2007; runoff data from U.S. Geological Survey, 2006b). National Oceanic and Atmospheric Administration, 2007a, Effects of Low Streamflows in Kansas, National Climatic Data Center: Information available on Web, Water Years 2000–06 accessed May 9, 2007, at http://www1.ncdc.noaa.gov/pub/data/ cirs/drd964x.pcp.txt • Record-low water levels at Milford Lake near Junction National Oceanic and Atmospheric Administration, 2007b, Palmer City (site 6, fig. 1, table 2) and Perry Lake near Perry Drought Severity Index by division: Information available on (site 20, fig. 1, table 2) in water year 2003. Water eleva- Web, accessed February 19, 2007, at http://www1.ncdc.noaa. gov/pub/data/cirs/drd964x.pdsi.txt tions at Kanopolis Lake near Kanopolis (site 9, fig. 1, table 2) were below conservation pool all of water year Paulson, R.W., Chase, E.B., Roberts, R.S., and Moody, D.W., 2006. 1991, National water summary 1988–89—hydrologic events and floods and droughts: U.S. Geological Survey Water-Supply • A water emergency in Salina because of extreme low Paper, 2175, p. 289–294. streamflows in Smoky Hill River in July, 2006. Salina’s PRISM Group, Annual precipitation data, Information available principle water supply is from alluvial wells in connec- on Web, accessed November 14, 2007 at http://www.prism. tion to the Smoky Hill River. oregonstate.edu/ • A record number of minimum desirable streamflow Putnam, 2003, Trends in suspended-sediment concentration at (MDS) violations (State administers MDS at Kansas selected stream sites in Kansas, 1970–2002: U.S. Geological streams to maintain adequate streamflow for habitat Survey Water Resources Investigations Report 03–4150, 36 p. and water supply; Katie Tietsort, Kansas Department Putnam, J.E., and Schneider, D.R., 2003, Water resources data, of Agriculture, Division of Water Resources, written Kansas, water year 2003: U.S. Geological Survey Water-Data comm., 2007). This reflects degradation of stream habi- Report KS–03–1, 641 p. tat as well as the curtailment of many water users. Seaber, P.R., Kapinos, F.P., and Knapp, G.L., 1987, Hydrologic unit maps: U.S. Geological Survey Water-Supply Paper 2294, 63 p. What Does the Future Hold? U.S. Army Corps of Engineers, 2007, National inventory of dams: Information available on Web, accessed September 7, 2007 at Streamflow conditions during future droughts in Kansas http://crunch.tec.army.mil/nidpublic/webpages/nid.cfm may look more like those that occurred during 2000–06 on the U.S. Geological Survey, 2006a, Past streamflow conditions: Republican, Saline, Solomon, and Smoky Hill Rivers than those Information available on Web, accessed November 17, 2006, at during the 1950s largely because of the factors noted earlier http://water.usgs.gov/waterwatch/?m=statesum&r=ks in this paper. In fact, if a sustained period of below-normal U.S. Geological Survey, 2006b, Surface-water annual statistics precipitation occurs, comparable to the 1950s, future hydro- for Kansas: Information available on Web, accessed May 10, logical droughts in Kansas probably will be much more severe 2007, at http://waterdata.usgs.gov/ks/nwis/annual/?referred_ than any previous droughts on record, and streamflows will be module=sw adversely affected. If climate change occurs and results in even U.S. Geological Survey, 2006c, Annual water data reports, more severe precipitation deficits during droughts, streamflow Information available on the Web at http://wdr.water.usgs.gov/ probably will be even more severe; therefore, State programs wy2006/pdfs/06865000.2006.pdf initiated to manage water supplies for users during drought that are based on the 1950s drought may need to be modified, U.S. Geological Survey, 2007a, 2006, USGS Streamflow infor- or new programs may need to be developed to better reflect mation on drought, Information available on Web, accessed the lower streamflows that may occur in future droughts with August 2006 at http://water.usgs.gov/waterwatch/?m=pa07d_ dry&r=ks&w=pa07d_dwc%2Cplot similar, or even less amounts, of precipitation that occurred in the 1950s drought. More study also is needed to understand the U.S. Geological Survey, 2007b, Streamflow information on historic change in runoff ratio and the factors that contribute to drought-summary plot of below normal streamflow: Information this change. available on Web, accessed May, 2007, at http://water.usgs.gov/ waterwatch/?m=pa07d_dry&r=ks&w=pa07d_dry%2Cplot

References Publishing support provided by: Rolla Publishing Service Center Hayes, M.J., 2002, Drought indices, in What is Drought?: Lincoln, For additional information visit the USGS Web site at: Nebraska, National Drought Mitigation Center, University of http://ks.water.usgs.gov/Kansas/waterwatch/drought/index. Nebraska, Information available on Web, accessed January 16, shtml 2007, at http://drought.unl.edu/whatis/indices.htm or contact: Nace, R.L., Pluhowski, E.J., 1965, Drought of the 1950s with Director, USGS Kansas Water Science Center special reference to the mid-continent: U.S. Geological Survey 4821 Quail Crest Place Water-Supply Paper 1804, 88p. Lawrence, Kansas 66049–3839 785–842–9909 or E-mail: dc_ks.usgs.gov