Estimating the magnitude and frequency of large floods and droughts and their effect on people also is difficult. One of the primary missions of the U.S. Geological Survey is to operate a streamflow- gaging-station network to monitor the Nation's water resources and to evaluate streamflow extremes. Estimates of the frequency of floods, droughts, and long-term streamflow variability from short-term (generally much less than 100 years) data records contain much uncertainty. Paleohydrologic techniques offer a way to lengthen a short-term data record and, therefore, to reduce the uncertainty in hydrologic analysis. Paleohydrology, as discussed in this article, is the study of the evidence of the movement of water and sediment in stream channels before the time of continuous (systematic) hydrologic records or direct measurements (Costa, 1987). Paleohydrologic data typically have been used to quantitatively reconstruct hydrologic variability for about the last 10,000 years. Beyond 10,000 years, quantitative paleohydrologic investigations often are hindered by limited evidence of channel changes, which reduces the accuracy of such paleohydrologic estimates. Also, in glaciated areas ice-may have reworked channel deposits and features.
This article briefly reviews paleohydrologic techniques used for analyzing floods, droughts, and hydrologic variability and discusses the application of these techniques in estimating long-term hydrologic records. Emphasis of the review is on those techniques most commonly used for reconstructing stream discharge and for dating floods and droughts. The references cited provide sources for additional information about paleohydrology; excellent reviews of the history of paleohydrology are provided by Costa (1987) and Patton (1987).
Figure 1. Sources of data used to reconstruct climatic or hydrologic conditions. (Modified from Liebscher, 1987). Click to view full-size image.
Direct data can be systematic (measured) or historical. Systematic data, such as streamflow records obtained at gaging stations, have been collected for about the last 100 years as part of scientific investigations. Historical data are recorded episodic observations of streamflow conditions, floods, and droughts that were made before systematic data were collected. In the United States, historical data typically are available for 100 to 200 years (Thomas, 1987). In Egypt and China, historical records are available for several thousands of years (Baker, 1987).
Indirect data, such as pollen, sediment, and tree-ring records, are examples of what is known as proxy data. Each type of proxy data has different problems related to its accuracy. Scientists use proxy data to extend climatic and hydrologic records. For example, tree-ring data have been used to reconstruct past precipitation and temperature for several hundred to thousands of years; deep-sea sediment cores have been used to reconstruct long-term, global temperature fluctuations for thousands to hundreds of thousands of years.
Paleohydrologic analysis uses many types of proxy data. Evidence of historic and prehistoric floods commonly is preserved in stream channels as distinctive sedimentologic deposits or landforms and also can be preserved as botanical evidence. The interpretation of this evidence provides important supplemental information about the spatial occurrence, magnitude, age, and frequency of floods, droughts, and hydrologic variability.
Figure 2. Examples of long-term temperature variability and sea-level changes based on several types of paleohydrological evidence. A, principal trends in average temperature during the last 100,000 years in northwestern Europe. B, Estimated sea-level changes during the past 150,000 years, which primarily reflect storage and release of water from glaciers. (Modified from (A) Knighton, 1984, (B) Moore, 1982). Click to view full-size image.
Climatic change involves changes in the solar-energy regime of a given region that affect the hydrologic cycle. The adjustments of the hydrologic cycle to long-term variability and climatic change also are recorded in surficial sediment deposits and landforms. Variations in lake and ocean levels provide an indication of climatic variability, and ocean-bottom sediments have been analyzed to reconstruct temperatures for as much as about 100,000 years ago (Knighton, 1984). Examples of broad averages of long-term temperature variability and sea-level changes based on several types of paleohydrologic evidence are shown in figure 2. Data in these graphs reflect substantial variations in climate and indicate that the present state of the hydrologic cycle is transient, when considered in the context of millennia. Because climate affects water-resources availability, long-term data are needed to assess the effects of climatic change on water resources. Consequently, there is a need to better determine past streamflow to assess current values.
Recent advances have been made in dating techniques, in paleodischarge estimation, and in the determination of the recurrence interval of floods (Baker, 1987; Baker and others, 1988). Many techniques are available in the scientific disciplines of sedimentology, geomorphology, hydraulics, and botany to extend hydrologic records, and the paleohydrologic techniques expand on those techniques to analyze floods, droughts, and hydrologic variability.
Figure 3. An ideal channel for studying slack-water deposits--the Escalante River in Utah. The person to the left is standing on a typical sequence of slack-water deposits that were deposited where the flow velocity decreased in the canyon of the Escalante River. (Source: Robert H. Webb, USGS).
Absolute dating and assignment of a specific age to a sediment sample using radiocarbon dating (Baker, 1987) are derived from laboratory determination of the decay of radioactive carbon-14 to stable carbon-12 in a sample of organic material. Materials used for radiocarbon dating are wood, charcoal, leaves, humus in soils, and other organic material. Samples of material are collected from slack-water deposits, gravel bars, and other depositional features. Recent advances in radiocarbon dating, such as the use of a tandem accelerator mass spectrometer, allow for dating extremely small samples with great precision (Baker, 1987). Samples having an age of 10,000 years generally can be dated with an uncertainty of less than 100 years (Stedinger and Baker, 1987).
Several different types of botanic evidence for determining the age or chronology of floods and droughts occur on flood plains (Sigafoos, 1964; Fritts, 1976; Hupp, 1987). Long-term annual tree-ring chronologies provide information for precisely dating floods and droughts, in some cases to the nearest year. An accurate tree-ring technique that is used to determine dates of floods is the analysis of tree-ring increments or wedge cuttings through scars. Analysis of a scar location in the tree rings can determine the year of flooding. Stewart and LaMarche (1967) provided documentation of the dates of flooding resulting from their investigations of the severe flood of 1964 in northern California. Their investigations in Coffee Creek indicated that many 200- to 400-year-old trees that had survived lesser floods were toppled during the flood of 1964. Costa (1978) indicated that the age of trees growing on flood-deposited sediment or in flood-scoured areas provides a minimum age since the most recent flooding.
Figure 4. Climatic effect on magnitudes of floods of a given probability, Mississippi River in Minnesota, 1867-1980. (Modified from Knox, 1984). Click to view full-size image.
Knox (1984) suggested that climate has affected flood magnitude in the Mississippi River in Minnesota from 1867 through 1980. The annual flood series was partitioned into four time (climatic) intervals. The intervals were chosen because independent studies of climate indicated that the boundaries of these intervals represent approximate dates of changes in characteristics of large-scale atmospheric circulation patterns that control the extent of air masses and the position of storm tracks. Flood-frequency analysis indicated that the magnitude of floods of a given probability varied substantially between the four climatically defined intervals (fig. 4). The first and last intervals were cooler, more moist, and prone to flooding. The greatest influence is on the magnitude of larger floods. Because flood-plain-management policies are based on estimated magnitudes of large floods, generally the 100-year flood, potential effects of climatic variability might need to be considered.
The Colorado River, which is a major source of surface water for much of the southwestern United States, traverses some of the most arid land in the country. The water in the Colorado River is over allocated and is regulated legally by the Colorado River Compact, which allocates the water to the many users in the United States and Mexico. For the 35 years from 1896 through 1930, the period on which the compact is based, the average discharge of the Colorado River was about 23,500 ft3/s (17 million acre-feet per year). However, from 1931 through 1965, the average discharge was only about 13 million acre-feet per year. A long-term (450 years) streamflow record reconstructed from tree rings (fig. 5) indicated the average discharge to be about 18,600 ft3/s (13.5 million acre-feet per year), much less than the 24,200 ft3/s (17.5 million acre-feet per year) on which the compact is based (Stockton and Boggess, 1983).
Figure 5. Estimated Colorado River discharge for 450 years (before 1960) at Lees Ferry, Ariz., based on reconstruction from tree-rings. (Modified from Stockton and Boggess, 1983). Click to view full-size image.
The 450-year streamflow record reconstructed from tree rings indicated that the period on which the compact is based contained the longest series of high-flow years during the entire 450-year reconstructed streamflow records. Also, droughts from 1564 through 1600 and from 1868 through 1892 were of longer duration and greater magnitude than for any period of the gaged record. In the Upper Colorado River Basin, climatic changes related to the possible "greenhouse effect" (slightly warmer temperature and less precipitation) may decrease streamflow by 35 percent (Stockton and Boggess, 1982). These streamflow data reconstructed from the analysis of tree rings indicate that climatic variability might reduce the future availability of water from the Colorado River to support continued development of the Southwest.
Although many paleohydrologic techniques are available, research can improve the understanding of physical processes of floods and droughts, paleohydrologic techniques, the understanding of climatic and hydrologic variability, and statistical procedures to better use historical and paleohydrologic data. The results of this research may reduce the uncertainty of hydrologic modeling, which in turn will decrease the uncertainty of water-supply estimates and flood estimates in water-resources planning.
Baker, V.R., 1983, Large-scale fluvial geomorphology, in Gregory, R.J., ed., Background to paleohydrology-A perspective: New York, John Wiley, p. 453-478.
____1987, Paleoflood hydrology and extraordinary flood events: Journal of Hydrology, v. 96, no. 1-4, p. 79-99.
Baker, V.R., Kochel, R.C., and Patton, P.C., eds., 1988, Flood geomorphology: New York, John Wiley, 503 p.
Baker, V.R., Kochel, R.C., Patton, P.C., and Pickup, G., 1983, Paleohydrologic analysis of Holocene flood slack-water sediments, in International Conference on Fluvial Sedimentology, 2d, Glasgow, Scotland, 1981, Proceedings: Keele, England, Blackwell, p. 229-239.
Blasing, T.J., and Fritts, H.C., 1976, Reconstructing past climatic anomalies in the north Pacific and western North America from tree-ring data: Quaternary Research, v. 6, no. 4, p. 563-579.
Chow, V.T., 1959, Open-channel hydraulics: New York, McGraw-Hill, 680 p. Costa, J.E., 1978, Holocene stratigraphy in flood-frequency analysis: Water Resources Research, v. 14, no. 4, p. 626-632.
____1983, Paleohydraulic reconstruction of flash-flood peaks from boulder deposits in the Colorado Front Range: Geological Society of America Bulletin, v. 94, no. 8, p. 986-1004.
____1987, A history of paleoflood hydrology in the United States, 1800-1970, in Landa, E.R., and Ince, Simon, eds., The history of hydrology: Washington, D.C., American Geophysical Union, p. 49-53.
Duvik, D.N., and Blasing, T.J., 1981, A dendroclimatic reconstruction of annual precipitation amounts in lowa since 1860: Water Resources Research, v. 17, no. 4, p. 1183-1189.
Fritts, H.C., 1976, Tree rings and climate: London, Academic Press, 567 p.
Gardner, T.W., 1983, Paleohydrology and paleomorphology of a carboniferous, meandering, fluvial sandstone: Journal of Sedimentary Petrology, v. 53, no. 3, p. 991-1005.
Gregory, K.J., ed., 1983, Background to palaeohydrology—A perspective: New York, John Wiley, 486 p.
Harrison, S.S., and Reid, J.R., 1967, A flood-frequency graph based on tree-scar data: North Dakota Academy of Science Proceedings, v. 21, p. 23-33.
Hupp, C.R., 1987, Botanical evidence of floods and paleoflood history, in Singh, V.P., ed., Regional flood frequency analysis-International Symposium on Flood Frequency and Risk Analysis, Baton Rouge, La., May 1986, Proceedings: Dordrecht, Holland, D. Reidel, p. 355-369.
Jarrett, R.D., 1988, Hydroclimatic data errors and their effects on the perception of climate change, in Pielke R.A., and Kittel, T.G.F., eds., Monitoring Climate for the Effects of Increasing Greenhouse Gas Concentrations, Pingree Park, Colo., 1987, Proceedings: Fort Collins, Colorado State University, Cooperative Institute for Research in the Atmosphere, p. 149-158.
Jarrett, R.D., and Costa, J.E., 1988, Evaluation of the flood hydrology in the Colorado Front Range using precipitation, streamflow, and paleoflood data: U.S. Geological Survey Water-Resources Investigations Report 87-4117, 37 p.
Jarrett, R.D., and Malde, H.E., 1987, Paleodischarge of the late Pleistocene Bonneville Flood, Snake River, Idaho, computed from new evidence: Geological Society of America Bulletin, v. 99, no. 1, p. 127-134.
Knighton, David, 1984, Fluvial forms and processes: London, Edward Arnold, 218 p.
Knox, J.C., 1984, Fluvial response to small scale climate changes, in Costa, J.E., and Fisher, P.J., eds., Developments and applications of geomorphology: Berlin, Springer-Verlag, p. 318-342.
Kochel, R.C., and Baker, V.R., 1982, Paleoflood hydrology: Science, v. 215, no. 4531, p. 353-361.
Liebscher, H.J., 1987, Paleohydrologic studies using proxy data and observations, in Solomon, S.I., Beran, M., and Hogg, W., eds., The influence of climate change and climatic variability on the hydrologic regime and water resources—International Association of Hydrological Sciences [IAHS] Symposium, Vancouver, Canada, 1987, Proceedings: Wallingford, England, International Association of Hydrological Sciences Press, IAHS-AISH Publication 168, p. 111-121.
Meier, M.F., 1986, Snow, ice, and climate—Their contribution to water supply, in U.S. Geological Survey, National water summary 1985-Hydrologic events and surface-water resources: U.S. Geological Survey Water-Supply Paper 2300, p. 69-82.
Meko, D.M., 1982, Drought history in the western Great Plains from tree rings, in Johnson, A.I., and-Clark, R.A., eds., Intemational Symposium on Hydrometeorology, Denver, Colo., 1982, Proceedings: Bethesda, Md., American Water Resources Association, p. 321-326.
Moore, W.S., 1982, Late Pleistocene sea-level history, in Ivanovich, M., and Harmon, R.S., eds., Uranium series disequilibrium-Applications to environmental problems: Oxford, England, Clarendon Press, p. 41.
Patton, P.C., 1987, Measuring the rivers of the past-A history of fluvial paleohydrology, in Landa, E.R., and Ince, Simon, eds., The history of hydrology: Washington, D.C., American Geophysical Union, p. 55-67.
Patton, P.C., and Baker, V.R., 1977, Geomorphic response of central Texas stream channels to catastrophic rainfall and runoff, in Doehring, D.O., ed., Geomorphology in arid regions: Fort Collins, Colo., Doehring, p. 189-217.
Patton, P.C., and Dibble, D.S., 1982, Archeologic and geomorphic evidence for the paleohydrologic record of the Pecos River in west Texas: American Journal of Science, v. 282, no. 2, p. 97-121. Puckett, L.J., 1981, A drought index for northern Virginia: U.S. Geological Survey Water-Supply Paper 2080, 39 p.
Schumm, S.A., 1977, The fluvial system: New York, John Wiley, 338 p.
Sigafoos, R.S., 1964, Botanical evidence of floods and floodplain deposition: U.S. Geological Survey Professional Paper 485-A, 35 p.
Solomon, S.I., Beran, M., and Hogg, W., eds., 1987, The influence of climate change and climatic variability on the hydrologic regime and water resources-International Association of Hydrological Sciences [IAHS] Symposium, Vancouver, Canada, 1987, Proceedings: Wallingford, England, International Association of Hydrological Sciences Press, IAHS-AISH Publication 168, 640 p.
Stedinger, J.R., and Baker, V.R., 1987, Surface water hydrology-Historical and paleoflood information: Reviews of Geophysics, v. 25, no. 2, p. 119-124.
Stedinger, J.R., and Cohn, T.A., 1986, Flood frequency analysis with historical and paleoflood information: Water Resources Research, v. 22, no. 5, p. 785-793.
Stewart, J.H., and LaMarche, V.C., 1967, Erosion and deposition produced by the flood of December 1964 on Coffee Creek, Trinity County, California: U.S. Geological Survey Professional Paper 422-K, 22 p.
Stockton, C.W., 1975, Long-term streamflow records reconstructed from tree rings: Tucson, University of Arizona Press, Laboratory Tree-Ring Research Paper 5, 111 p.
Stockton, C.W., and Boggess, W.R., 1982, Climatic variability and hydrologic processes—An assessment for the southwestern United States, in Johnson, A.I., and Clark, R.A., eds., International Symposium on Hydrometeorology, Denver, Colo., 1982, Proceedings: Bethesda, Md., American Water Resources Association, p. 317-320.
____1983, Tree-ring data-Valuable tool for reconstructing annual and seasonal streamflow and determining long-term trends, in Transportation Research Board, National Academy of Sciences, Improving estimates from flood studies: Transportation Research Record 922, p. 10-17.
Stockton, C.W., and Fritts, H.C., 1973, Long-term reconstruction of water level changes for Lake Athabasca: Water Resources Bulletin, v. 9, no. 5, p. 1006-1027.
Stokes, W.L., Judson, Sheldon, and Picard, M.D., 1978, Introduction to geology (2nd ed.): Englewood Cliffs, N.J., Prentice-Hall, 656 p.
Thomas, W.O., Jr., 1987, The role of flood-frequency analysis in the U.S. Geological Survey, in Singh, V.P., ed., Application of frequency and risk in water resources—International Symposium on Flood Frequency and Risk Analysis, Baton Rouge, La., May 1986, Proceedings: Dordrecht, Holland, D. Reidel, p. 463-484.
U.S. Army Corps of Engineers, 1989, Annual flood damage report to Congress, fiscal year 1989—prepared in cooperation with the National Weather Service: Washington, D.C., U.S. Army Corps of Engineers DAEN-CHW-W.
Wahl, K.L., 1984, Evolution of the use of channel cross-section properties for estimating streamflow characteristics, in Meyer, E.L., ed., Selected papers in the hydrologic sciences: U.S. Geological Survey Water-Supply Paper 2262, p. 53-66.
Webb, R.H., and Baker, V.R., 1987, Changes in hydrologic conditions related to large floods on the Escalante River, south-central Utah, in Singh, V.P., ed., Regional flood frequency analysis-International Symposium on Flood Frequency and Risk Analysis, Baton Rouge, La., May 1986, Proceedings: Dordrecht, Holland, D. Reidel, p. 309-323.
Williams, G.P., 1983, Paleohydrological methods and some examples from Swedish fluvial environments, pt. I-Cobble and boulder deposits: Geografiska Annaler, ser. A [Physical Geography], v. 65A, no. 34, p. 227-243.
____1984, Paleohydrologic equations for rivers, in Costa, J.E., and Fisher, P.J., eds., Developments and applications of geomorphology: Berlin, Springer- Verlag, p. 343-367.
____1986, River meanders and channel size: Journal of Hydrology, v. 88, nos. 1-2, p. 147-164.
Winter, T.C., and Wright, H.E., Jr., 1977, Paleohydrologic phenomena recorded by lake sediments: Eos, American Geophysical Union, v. 58, no. 3, p. 188-196.