Hirschboeck, K.K., 1991, Climate and floods, in Paulson, R.W., Chase, E.B., Roberts, R.S., and Moody, D.W., Compilers, National Water Summary 1988-89--Hydrologic Events and Floods and Droughts: U.S. Geological Survey Water-Supply Paper 2375, p. 99-104.
With additional information from:
House, P.K., and Hirschboeck, K.K., (in press), Hydroclimatological and paleohydrological context of extreme winter flooding in Arizona, 1993: Geological Society of America Special Paper.
Ely, L.L., Enzel, Y., Baker, V.R., and Cayan, D.R., 1993, A 5000-year record of extreme floods and climate change in the southwestern United States: Science, v. 62, p. 410-412.
Trenberth, K.E., and Hoar, T.J., 1996, The 1990-1995 El Niño-Southern Oscillation event longest on record: Geophysical Research Letters, v. 23, No. 1, p. 57-60.
Floods are caused by weather phenomena and events that deliver more precipitation to a drainage basin than can be readily absorbed or stored within the basin. The kinds of weather phenomena and events that cause floods include intense convective thunderstorms, tropical storms and hurricanes, cyclones and frontal passages, and rapid snowmelt. These individual meteorological processes are part of a larger climatic framework that determines (1) the seasonal availability and large-scale delivery pathways of atmospheric moisture, (2) the seasonal frequency, typical locations, and degree of persistence of the weather phenomena that release the delivered moisture, and (3) the seasonal variation of climate-related, land-surface conditions that affect flood runoff, such as antecedent soil moisture or snow cover.
Although the southwestern U.S. is noted for its general dryness, floods are not uncommon. Summer thunderstorms, producing flash floods that suddenly fill dry washes and overwhelm whatever is in their way are famous.
Figure 1. Precipitable water vapor in the atmosphere over the conterminous United States. Upper maps, 2-6 miles above sea level; lower maps, surface to 6 miles above sea level. A, average annual, B, average July. (Data from Reitan, 1960)
The greatest influx of precipitable water vapor across the entire conterminous United States occurs in July (fig. 1B). During July, the maximum available precipitable water vapor is present in the atmosphere over the southeastern United States, especially over the Gulf Coast States. Even at altitudes higher than 2 miles, the precipitable water vapor in the atmosphere is greatest in July, and at these altitudes, a northward influx of moisture into the western United States is evident. Only at altitudes between 2 and 6 miles above sea level over the Far Northern and Western States does the atmosphere contain less than 0.4 inch of precipitable water vapor during July.
LARGE-SCALE, MOISTURE-DELIVERY PATHWAYS
The primary large-scale, moisture-delivery pathways along which low-level precipitable water vapor enters the conterminous United States are determined by the direction of average surface winds at different times during the year. These pathways can be depicted for each season by using simplified streamlines of the average surface winds over North America (fig. 2). In figure 2, generalized pathways are shaded to reflect the average precipitable water vapor in the lower atmosphere over different areas of the continent during each season. At any given time, the dominant pathways over the conterminous United States originate from three distinct air-mass source regions--the Pacific Ocean, the Atlantic Ocean and Gulf of Mexico, and the Arctic region (Bryson and Hare, 1974; Wendland and Bryson, 1981). In addition, in fall and winter, some pathways originate from air masses located over the north-central and eastern United States (Bryson and Hare, 1974; Wendland and Bryson, 1981).
Figure 2. Large-scale moisture delivery pathways over North America in 4 midesason months. A, January, B, April, C, July, D, October. (Pathways data from Bryson and Hare, 1974; precipitable water vapor data from Reitan, 1960)
Regionally, the dominant air masses and moisture-delivery pathways shift as the seasons progress. These shifts are important controls of the seasonality of average monthly precipitation and streamflow in the conterminous United States. These shifts also determine the regional flooding regimes of rivers by affecting the timing and magnitude of large influxes of precipitable water vapor that are necessary for intense or prolonged storms.
Moisture Delivery from the Pacific Ocean
A belt of strong westerly winds delivers moisture-laden air from the Pacific Ocean into North America along several pathways that shift with the seasons-as far north as about 60° N. latitude in summer and as far south as about 35° N. latitude in winter (fig. 2). Penetration of this Pacific air into the interior of the continent by the westerly winds is impeded at lower atmospheric levels by the mountain ranges along the west coast of the continent. Entry into the interior is possible where the westerly winds are strongest, between about 45° and 50° N. latitude, and where elevations of the mountain ranges are low (Bryson and Hare, 1974). Although air from the Pacific Ocean generally contains substantial precipitable water vapor before the air penetrates deep into the interior, the moisture in the air in the southern pathways, which affects much of the west coast of the United States, is not released abundantly as precipitation due to the proximity of the North Pacific subtropical anticyclone and the stabilizing effect of the cold California Current. These conditions are especially true in summer, which tends to be a dry season along the far West Coast. In the interior southwestern part of the continent, however, localized intense precipitation and flooding can occur during the summer when moisture from the warm, tropical parts of the Pacific Ocean is conveyed northward along the west coast of Mexico into parts of southern California and Arizona. In the middle latitudes of the Far West, major occurrences of intense precipitation and flooding take place in winter, when well-developed, extratropical, cyclonic storms and fronts are conveyed into the region by strong westerly winds that have shifted to more southerly pathways. In the northern latitudes of the West Coast and in Alaska, moisture delivery by pathways from the Pacific Ocean is most effective in late summer and fall.
Air from the Pacific Ocean that crosses the western mountain ranges and enters the north-central and eastern parts of the continent loses much of its precipitable water vapor and has a modified character that contrasts with the cold, dry air from the Arctic and the warm, moist subtropical air from the Atlantic Ocean and Gulf of Mexico. This modified air from the Pacific Ocean has a drying effect when it dominates during the fall, thereby decreasing the likelihood of flooding. During the winter, however, modified air from the Pacific Ocean occasionally can be the source of unseasonably warm temperatures and dry air, which result in a temporary thaw that produces significant snowmelt flooding in the Northern States.
Air that is conveyed across the tropical part of the Pacific Ocean by the northeasterly trade winds has a continuous effect on Hawaii. This State receives abundant precipitable water vapor in all months of the year, but most of it is concentrated at low levels in the atmosphere, below 5,000 to 8,000 feet above sea level. Above this altitude, the upper trade winds and high pressure from the North Pacific subtropical anticyclone typically produce an inversion that has a stable layer of dry air above. Light to moderate rain showers are common under these conditions, but occasionally extreme rainfall intensities and subsequent flooding can occur, especially during periods of frontal passage and storm activity in the winter and tropical cyclone activity in the summer and fall.
Convectional processes tend to have limited areal extent but cause intense precipitation that results in flash floods in small drainage basins. Frontal-convergence processes affect extensive areas, and the resulting precipitation is the predominant cause of floods in large drainage basins. Upper atmosphere perturbations can have a local or widespread effect on precipitation and flooding, depending on the small-scale or large-scale nature of the perturbation. Orographic uplifting also can have a local or widespread effect on precipitation and flooding, depending on the extent and orientation of the topography with respect to prevailing wind directions and the type of precipitation associated with the orographic uplift.
Convectional processes that are capable of releasing enough precipitation to cause floods include thermally induced thunderstorms, mesoscale convective systems, and tropical cyclones. In addition, convectional processes also can occur in conjunction with frontal convergence and orographic lifting. The characteristic feature of a strong convectional process is the towering thunderstorm cloud from which exceedingly intense precipitation of short duration can fall.
Extreme precipitation intensities of as much as 1.26 inches in 1 minute (Maryland), 3.9 inches in 14 minutes (Texas), 7 inches in 30 minutes (Ohio), and 12 inches in 42 minutes (Missouri) have been recorded in the United States in association with summer convectional processes (Costa, 1987). Precipitation of such great intensity is delivered at a rate faster than the rate at which it infiltrates the soil; hence, it contributes to surface runoff almost immediately, and commonly results in urban street flooding and flash flooding within small drainage basins. Most thunderstorm cells are of limited areal extent [4 to 40 mi2], whereas large, convectively induced and maintained thunderstorm clusters, called mesoscale convective complexes, can have areas of 30,000 to 80,000 mi2. The convectional processes associated with tropical storms and hurricanes can affect areas as large as 400,000 mi2. The seasonality and geographic distribution of each of these types of convectional processes are discussed in more detail in the following sections.
Intense precipitation from thunderstorms is the major cause of flash floods in small drainage basins and is an important component of many instances of widespread, severe flooding. Thunderstorms occur in all 50 States, but their frequency and spatial distribution vary from season to season. The major requirements for severe-thunderstorm development are warm, moist, unstable air and an atmospheric environment that supports low-elevation cloud bases and instability aloft so that clouds can form to great heights. Thunderstorms can develop in homogeneous warm, moist air masses or along fronts. In the former, instability is initiated by local heating, topographic effects, or upper atmosphere perturbations, whereas in the latter, instability is initiated by marked density contrasts between warm and cold air masses.
The average number of days during which thunderstorms develop in the United States varies from season to season (fig. 3), and there is a general relation between the number and geographic distribution of these days and the volume of precipitable water vapor in the lower atmosphere (figs. 1 and 2). Areas where there is an exception to this relation are along the west coast of the conterminous United States and Hawaii, where thunderstorms are infrequent despite an abundance of moist air from the Pacific Ocean. Along the West Coast, the cold California Current and the North Pacific subtropical anticyclone have a stabilizing effect on the atmosphere, whereas in Hawaii, an inversion in the trade winds having a stable, dry layer aloft suppresses thunderstorm development. Nevertheless, when thunderstorms do develop in these areas, the intense precipitation from these storms can cause major flooding, especially when atmospheric uplift and instability are enhanced by orographic effects.
Figure 3. Average number of days per season during which thunderstorms developed in the U.S., 1951-75.
Elsewhere in the United States, the number and geographic distribution of days during which thunderstorms develop varies from a minimum number of days during the winter, when thunderstorm activity is limited to the Gulf Coast region, to a maximum number of days during the summer, when thunderstorm activity occurs throughout most of the country (fig. 3A, C). The regions most prone to thunderstorm development are the Gulf Coast (all seasons), the Great Plains and the eastern slopes of the mountains just to the west of the Great Plains (spring, fall, and especially summer), and the southwestern deserts and mountains (summer). The large number of days during which thunderstorms develop in the Great Plains region is related to a variety of mechanisms, including frontal convergence, upper atmosphere perturbations, and mesoscale convective systems (discussed in the next section). In the mountainous parts of the western and southwestern regions, the forced up-slope movement of moist air during the summer greatly enhances thunderstorm development.
Despite their propensity for producing intense precipitation, most typical thunderstorms do not yield sufficient precipitation to cause major floods. A single storm usually is small, of short duration, and is characterized by several convective cells that develop and decay during an hour or so. Floods are most likely to be caused by precipitation from severe, long-duration thunderstorms that develop to great heights-- at times penetrating through an inversion layer to do so. These thunderstorms must have an internal structure that can withstand the dispersing effect of strong winds in the upper atmosphere. Development to the severe stage can be in the form of a single "super-cell" thunderstorm, a row of thunderstorms that forms a squall line, or a multicell cluster of developing and decaying thunderstorms (Eagleman, 1983). Due to the limited areal extent of precipitation from single, super-cell thunderstorms or small, multicell thunderstorms, flooding caused by the precipitation is localized, but it can be extreme. This condition is especially true when the thunderstorm remains quasi-stationary and is centered directly over a drainage basin or its headwaters or when a slow-moving thunderstorm moves along a path from the headwaters area to the discharge area of a linear drainage basin, thus allowing intense precipitation and resulting runoff to become superimposed on flood waters coming from upstream. On September 14, 1944, in a small drainage basin in Nevada (22.9 mi2), the latter situation resulted in one of the largest known flash floods for a drainage basin of this size (Glancy and Harmsen, 1975; Costa, 1987). In this instance, and in numerous other instances of major localized flooding in the United States that results from precipitation from thunderstorms, orographic lifting of warm, moist air has been an important component contributing to the severity of the thunderstorms.
Tropical cyclones are the largest atmospheric features that are induced and maintained primarily by convectional processes. Precipitation from these tropical low-pressure systems, which have diameters of 60 to 600 miles when fully developed, can cause major floods. Easterly waves, tropical depressions, tropical storms, and hurricanes each represent different stages of disturbed tropical weather associated with tropical cyclones. Those tropical cyclones that affect the conterminous United States originate in the western part of the North Atlantic Ocean, the Gulf of Mexico and Caribbean Sea, and the eastern part of the North Pacific Ocean (fig. 4). The point of initial genesis of these cyclones and their season of occurrence are affected principally by the presence of sea-surface temperatures of about 79 degrees Fahrenheit or warmer (Gray, 1979). The official tropical-cyclone season extends from June through October, but the greatest frequency of these cyclones is in late summer and early fall, when sea-surface temperatures tend to be warmest.
Figure 4. Tropical cyclones and their effect on flooding in the conterminous United States/
The effects of tropical cyclones on local and regional precipitation totals can be substantial. Along the coasts of the Atlantic Ocean and the Gulf of Mexico, precipitation from tropical cyclones contributed as much as 15 percent of the total precipitation from June through October during 1931-60 (Cry, 1967). Most of this precipitation occurred in the months of August and September, when 25 to 45 percent of the total monthly precipitation along the Gulf and Atlantic Coasts was contributed by tropical cyclones (Cry, 1967). In the southwestern United States, tropical cyclones originating in the eastern part of the North Pacific Ocean tend to have an even greater relative affect on precipitation totals during June through October because normal summer and fall precipitation totals are much less in this part of the country, which is more arid than the eastern United States. Several of the greatest precipitation totals ever recorded for individual storms in the southwestern United States have been associated with tropical cyclones, and these precipitation totals have helped to establish all-time records of total monthly precipitation at a number of stations scattered throughout the region (Smith, 1986).
Although floods caused by precipitation from tropical cyclones are infrequent, these floods can represent the largest recorded peaks in a flood time series; hence, these floods are important contributors to estimates of the 100-year and probable-maximum floods. Tropical cyclones can have an effect on flooding in three different ways:
1. As a cyclone approaches land, storm-surge waves and high water levels produced by strong onshore winds can cause severe coastal flooding and shoreline erosion. Extremely damaging coastal floods caused by such waves and water levels occurred during Hurricane Camille in August 1969, when storm surges of 15 to 22 feet above normal sea level were recorded at various locations along the Gulf Coast of Louisiana and Mississippi (Wilson and Hudson, 1969; U.S. Army Corps of Engineers, 1970).
2. After a cyclone has reached land, individual thunderstorms embedded within spiraling cloud bands of the cyclone can produce intense, localized precipitation that causes flash flooding in urban areas and small drainage basins. At the same time, widespread, intense precipitation that is generated by atmospheric instability throughout the extent of the cyclone can cause major riverine flooding in large drainage basins that are located along the path of the cyclone--especially if the cyclone is slow moving, becomes stalled, or reverses direction. The devastating floods during Hurricane Agnes that affected drainage basins throughout the Mid-Atlantic States in late June and early July 1972 were of this type (Bailey and others, 1975).
3. Regardless of whether or not a cyclone reaches land, residual moisture fluxes from any stage of a cyclone--including the dissipating stage--can introduce abnormally large volumes of precipitable water vapor into areas where substantial surface heating, orographic lifting, or upper atmospheric instability can produce intense precipitation sufficient to cause flooding. This is one of the most common ways in which a distant or dissipating tropical cyclone can be a major factor affecting flooding in the interior of the continent. The residual precipitable water vapor from a dissipating tropical cyclone is especially likely to cause flooding when this water vapor is incorporated with that in a preexisting atmospheric disturbance that moves into the area, such as a cutoff low, a front, or an extratropical cyclone. The record-breaking floods in southern Arizona in September 1970 (Tropical Storm Norma) and in October 1983 (Tropical Storm Octave) were of this type, as were the floods in southwestern Texas in June 1954 (Hurricane Alice) (fig. 4B).
On the basis of the tracks of previous tropical cyclones, most of the southwestern, south-central, and eastern parts of the conterminous United States are susceptible to occasional extreme flooding from precipitation or storm surges produced by the direct or indirect effects of a tropical cyclone (fig. 4B). Hawaii also is vulnerable to severe coastal and riverine flooding by precipitation and storm surges from tropical cyclones originating in the Pacific Ocean; however, Hawaii is not located in a source region of tropical cyclones, and fully developed tropical cyclones are not common in this part of the Pacific Ocean.
Figure 5. Regional paleoflood chronology showing floods on rivers in Arizona and southern Utah in 200-year increments (each bar represents the increment that ended on the indicated year). An asterisk indicates that one of the largest floods in the record occurred in the increment (multiple events indicated by the number in parentheses). The values for the last two increments are shown above the bars, as the bars exceeded the graph scale. (Modified from Ely and others 1993)
Between June and October, hurricane-like tropical cyclones generated in the Pacific west of Mexico and central America can sweep inland and cause flooding in the southwestern U.S.. For central and southern Arizona, Ely and others (1993, Fig. 5) showed that floods related to tropical cyclones were associated with a high in the Aleutians and a low in the western U.S., a pattern similar to that for severe winter floods (compare Figs. 6A and 6B).
Figure 6. Atmospheric circulation during severe winter and tropical cyclone floods. (A) Composite anomaly map of 700 mbar heights (in meters) assicaiated with winter floods on rivers in central Arizona; based on 8 cases from 1947-88. (B) Same as (A) for tropical cyclones assiciated with floods on rivers in central and southern Arizona; based on 7 cases. (Modified from Ely and others 1993)|
LARGE-SCALE ATMOSPHERIC CONVERGENCE
In contrast to thermal-convectional processes within a single homogeneous air mass, the development of widespread precipitation from fronts and extratropical cyclones is dependent on the convergence of air masses of marked density contrast. Precipitation resulting from large-scale atmospheric convergence tends to have a greater geographic extent, less intensity, and a longer duration than does that resulting from thermal-convectional processes. The presence of converging air masses, however, can enhance atmospheric instability that induces localized thermal-convectional processes; hence, it is common for frontal activity to be characterized by widespread, moderate precipitation due to large-scale atmospheric convergence and localized, intense precipitation due to thunderstorms or squall lines in the vicinity of the front. When this occurs, large-scale riverine flooding and local flash flooding are possible.
Due to its middle-latitude location, most of the conterminous United States underlies zones of convergence between polar and tropical air masses; the location of these zones varies from season to season (fig. 2). Specific flood-causing processes that are associated with large-scale atmospheric convergence are extratropical cyclones and their associated fronts and wind patterns in the upper atmosphere that enhance convergence and instability near the land surface.
Extratropical Cyclones and Their Associated Fronts
Most major floods in large drainage basins in the conterminous United States are caused by precipitation from extratropical cyclones and their associated fronts. Systematic observations of the movement of extratropical cyclones and fronts across North America have identified a series of preferred tracks that define regions most susceptible to repeated passages of these cyclones and fronts (fig. 7). These preferred extratropical cyclone tracks shift to their southernmost trajectories in the winter and to their northernmost trajectories in the summer and have intermediate trajectories in the spring and fall.
Figure 7. Primary tracks of extratropical cyclones in North America for four midseason months based on frequency of extratropical cyclones during 1951-70. A, January, B, April, C, July, D, October. (Modified from Reitan and others, 1974 and Reed, 1960)
In the winter, virtually all States in the conterminous United States are, at least in part, susceptible to flooding caused by precipitation from extratropical cyclones and their associated fronts (fig. 7A). Extratropical cyclones from the Pacific Ocean associated with the Aleutian Low storm track successively enter the Northwestern States and bring rain to lower elevations and snow to higher elevations--some of which can contribute to snowmelt flooding during thaws. In the Northeastern States, substantial winter precipitation results from extratropical cyclones that originate in Alberta and Colorado. In the higher elevations of the Northeastern States, deep snow can accumulate, due in part to snow squalls generated by the thermal effects of the Great Lakes. This snow accumulation can result in flooding during thaws or periods of warmer rainfall. In the winter, East Coast States from Georgia to Maine are susceptible to flooding from intense precipitation and coastal storm surges produced by extratropical cyclones that track northeastward along the Atlantic seaboard. Also in the winter, Gulf Coast States experience flooding due to the frequent presence of fronts in the region. In this region, cold fronts in the winter commonly are accompanied by thunderstorms that can produce intense precipitation in the form of rain. In addition, stationary fronts tend to be alined with the Gulf Coast and commonly produce widespread, persistent rain that saturates the soil and increases the likelihood of flooding.
In the spring, extratropical cyclones are quite active, and their trajectories shift slightly northward (fig. 7B). Extratropical cyclones that traverse the Southwestern and South-Central States become numerous in late winter and early spring, and frontal passages associated with these cyclones can produce substantial rainfall or late-season snowfall from the mountainous regions of the West all the way to the Northeast Coast. The other extratropical-cyclone tracks of spring are similar to those of winter but are shifted slightly to the north. The density contrasts between converging air masses are much greater in the spring, however, and greater volumes of precipitable water vapor are available in spring than in winter. These factors combine with the presence of saturated soil, frozen ground, and rapid snowmelt to make spring the season that has the greatest flooding potential in many areas of the conterminous United States.
During the summer, most extratropical cyclonic activity shifts to the Northern States and Canada and to the northeastern seaboard (fig. 7C). Intense precipitation from convective thunderstorms associated with these cyclones and fronts can cause floods, especially when the precipitable water-vapor content of the atmosphere is unusually large and when fronts or low-pressure centers move slowly or become stalled.
In the fall, extratropical-cyclone tracks are similar to those of summer but are shifted slightly to the south (fig. 7D). Northern-latitude extratropical cyclones are quite active during this season, but because the moisture content of the atmosphere is less than in the summer and soil-moisture storage has been depleted, flooding is less common than in the summer. An exception occurs in the far Northwestern States and Alaska, where extratropical cyclonic activity in the Gulf of Alaska is at a maximum. Here, strong extratropical cyclones from the North Pacific Ocean can produce widespread precipitation that may become locally intensified through orographic effects. Storm surges also can cause flooding along low-lying areas of the Alaskan coast.
Precipitation-Enhancing, Upper Atmospheric Air Patterns
The development and movement of extratropical cyclones and their associated fronts are related to large-scale convergence and divergence of strong winds in the overlying upper atmosphere. These strong winds have circulation patterns that vary in response to pressure and temperature gradients in the atmosphere. The fastest upper atmospheric winds, called jet streams, tend to overlie marked atmospheric temperature and density discontinuities near the surface of the Earth, especially those discontinuities that occur along frontal boundaries. The general direction of movement over North America is from west to east, but a sinuous movement pattern is typical, characterized by wave-like southerly and northerly components of wind direction. This sinuous movement of air in the upper atmosphere forms troughs of low pressure and ridges of high pressure that normally progress eastward from day to day. When this progression is stalled or blocked, the trough-and-ridge positions can remain quasi-stationary for several days and cause extratropical cyclones and associated fronts near the surface of the Earth to also become similarly quasi-stationary (fig. 8A). Under these conditions, the potential for severe flooding, such as occurred in April 1983 in Louisiana (fig. 8A), is great because prolonged precipitation is possible along the stalled front and because intense convective showers can develop repeatedly along squall lines in the unstable air in the vicinity of the front. A related situation that is likely to cause floods is the frequent recurrence of precipitation resulting from the repeated development of a trough-and-ridge pattern in approximately the same location during an extended period of several weeks or months. The precipitation that caused the severe spring floods of 1973 in the Mississippi River basin was a result of these conditions (Chin and others 1975).
Figure 8. Precipitation-enhancing, upper atmospheric air patterns over various parts of North America. A, Quasi-stationary trough-and-ridge patterns. B, Development and dissipation of a cutoff low. C, Movement of small-scale, short-wave troughs of low pressure through large-scale, upper atmospheric air patterns. D, Multiple meteorologic features interacting with upper atmospheric air patterns. (Modified from Muller and Faiers, 1984; Hansen and others, 1988; Maddox and others, 1980; Hansen and Schwarz, 1981.)
Another upper atmospheric air pattern that can produce flood-causing precipitation is the cutoff low (fig. 8B). This low is an upper atmospheric low-pressure system that originates as part of a trough but then becomes displaced to the south and is cut off from the main west-to-east component of wind direction. Cutoff lows may persist for several days in approximately the same location, especially if they are flanked by blocking ridges of high pressure to the west or east. When stalled in this manner, cutoff lows can induce atmospheric instability and precipitation near the surface of the Earth and can slow the eastward progression of fronts. The cutoff low shown in figure 8B was associated with flooding in east-central Colorado.
Pulses of intense precipitation caused by the movement of small-scale, short-wave troughs of low pressure through the large-scale, upper atmospheric troughs or ridges also are potential producers of flood-causing precipitation (fig. 8C). Many severe flash floods in the Western and Central States have been associated with these short-wave troughs, including several of the largest rainfall-runoff floods ever recorded in the United States (Maddox and others, 1979, 1980; Hirschboeck, 1987a).
Commonly, a precipitation-enhancing upper-atmospheric air pattern will occur in conjunction with other precipitation-producing meteorological features, and an extreme or catastrophic flood will ensue. Several such floods have occurred in the Southwestern States when moisture from dissipating tropical cyclones, which originated in the Pacific Ocean, became incorporated into cutoff lows, small-scale, short-wave troughs of low pressure, or slow-moving fronts near the surface of the Earth (fig. 8D). During the 3 days shown in figure 8D, moisture from dissipating Tropical Storm Norma interacted with an upper atmospheric cutoff low and a cold front to produce substantial precipitation that caused flooding in central Arizona. In the Central and Eastern States, floods caused by precipitation from the interaction of similar, multiple meteorological features have occurred, such as the floods of August 1978 in central Texas that were initiated by precipitation from remnants of Tropical Storm Amelia and compounded by precipitation caused by other lower and upper atmospheric meteorological factors (Schroeder and others, I987). Another multiple-feature interaction that is typical of the eastern United States is the merging of a tropical cyclone with an upper atmospheric trough of low pressure and its subsequent transformation into an extratropical cyclone. This occurred during Hurricanes Agnes, Carla, and Betsy (fig. 4B).
In mountainous and hilly terrain, the release of moisture from the atmosphere by localized thermal convection, large-scale frontal convergence, or upper-air perturbations can be augmented by the process of orographic lifting. Orographic lifting is the forced ascent of an air current caused by its passage up and over a topographic barrier. Moist air that is lifted in this manner can cool sufficiently so that clouds will form and release some of the moisture as precipitation. Precipitation usually is greater at high elevations than on lowlands because of this process. Clouds and precipitation resulting from orographic lifting tend to be concentrated on the windward slopes of topographic barriers, so that windward slopes often are wet whereas the lee sides of slopes often are quite dry. Precipitation from orographic lifting can be locally intense, but it usually takes the form of moderate rain or snow that occurs regularly and persistently over specified areas and often yields extremely high totals of annual precipitation. When orographic lifting occurs in combination with other processes that release moisture in the atmosphere, substantial precipitation can result to produce a flood.
Orographically enhanced precipitation can occur wherever there are mountains or irregular terrain but is most common in regions where mountain ranges lie perpendicular to the direction of prevailing winds that transport moisture-laden air. In the eastern part of the conterminous United States, a moderate orographic effect occurs when air from the Atlantic Ocean or Gulf of Mexico is forced up the foothills and slopes of the mountainous terrain that extends from Tennessee to Maine. In the western pan of the conterminous United States, the greatest orographic enhancement of precipitation occurs in the mountain ranges that parallel the west coasts of Washington, Oregon, and northern California and that lie in the path of westerly winds from the Pacific Ocean. In Alaska, orographically enhanced precipitation occurs where air currents from the Pacific Ocean move onshore and up the slopes of coastal mountain ranges, especially along the southern coast and in the southeastern panhandle region of the State. In Hawaii, steep, windward slopes that are perpendicular to the prevailing northeasterly trade winds receive abundant precipitation, whereas the lee sides of these slopes are often quite dry.
Orographic lifting can be a factor in the occurrence of floods in several ways. Local flash floods are often the result of convective precipitation that has been orographically enhanced. The ascent of warm, moist, unstable air up the slopes of hills and mountains supports and enhances the development of severe thunderstorm cells that can yield substantial rainfall. Steep-sided valleys that are open to the inflow of moisture-laden winds can funnel moist air up the valley to the headwater area of a drainage basin. This topographically forced atmospheric convergence enhances instability in the air and promotes the development of clouds and heavy precipitation over the headwater area. In these instances, the terrain sometimes has an anchoring effect on a developing severe storm and causes the storm to remain quasi-stationary, while one thunderstorm cell after another is generated in roughly the same location. Precipitation from such a storm can be excessive and lead to severe flooding in small drainage basins. The destructive flash floods of June 9-10, 1972, in southwestern South Dakota (Schwarz and others, 1975) and July 31-August 1, 1976, in central Colorado (McCain and others, 1979) were floods caused by intense precipitation that was enhanced by orographic lifting.
Orographic lifting can be an important factor in the occurrence of floods even in regions that do not have major mountain ranges. When abundant moisture is present in the air and other flood-causing meteorological processes are active, orographic lifting over hilly terrain often combines with these processes and enhances the development of thunderstorms that produce excessive precipitation. The floods of August 1978 in central Texas were of this type (Schroeder and others, 1987).
Orographic lifting also has an influence on widespread, regional flooding. The persistence and regularity of orographically enhanced precipitation in certain areas can make these areas more susceptible to flooding because soils are more likely to stay close to saturation in such regions. Furthermore, many areas dominated by orographic lifting lie in the path of persistent storm tracks. In these regions, the passage of extratropical cyclones and their associated fronts can deliver widespread precipitation which, when locally enhanced and intensified by orographic lifting, can cause flooding in many drainage basins over a large area. The floods of December 1964-January 1965 in northern California were of this type (Waananen and others, 1971). Finally, winter precipitation in mountainous regions where orographic lifting is active usually results in the accumulation of extremely great snow depths at high altitudes. Spring snowmelt runoff from these areas can cause widespread flooding, especially during periods of rapid warming and when rain falling on snow accelerates the snowmelt process.
When a soil has a substantial infiltration capacity, it can absorb all but the most prolonged or intense precipitation, thereby limiting excessive surface runoff. Once the soil is saturated, however, even light to moderate precipitation will become surface runoff, and the potential for flooding increases. The volume of moisture held in a soil at any given time is controlled climatically by the difference between precipitation that infiltrates the soil and evapotranspiration that removes moisture from the soil. The relation between these two processes varies geographically and seasonally. Precipitation varies in response to the moisture-delivery pathways and meteorological processes discussed earlier. Evapotranspiration varies in response to available solar and heat energy and is greatest during the summer, when days are longest and temperatures are highest. Evapotranspiration is least during the winter, especially at northern latitudes, where days are short and temperatures are low.
Figure 9. effects of average monthly precipitation and soil moisture on susceptibility to flooding at San Diego, Calif., and Tucson, Ariz. Explanation: Blue bar, precipitation (inches); gray area, available soil-moisture storage capacity (in); pale blue area, soil moisture in storage (in); red ruled area, period of greatest susceptibility to flooding. (data from Mather, 1964)
In most areas of the United States, the volume of stored soil moisture is greatest during the late winter and early spring, at the end of the period of least evapotranspirarion (fig. 9). Conversely, due to substantial evapotranspiration in the summer, moisture stored in the soil is least during the late summer and early fall throughout much of the Nation. This relation is true even in regions that have substantial precipitation in the summer. [For additional information on evapotranspiration, see "Evapotranspiration and Droughts"]
These seasonal fluctuations in soil-moisture storage determine the susceptibility of an area to flooding at different times during the year (fig. 9). In most parts of the United States, the flooding potential from a storm of a given magnitude is greatest in the late winter and early spring and least in the late summer and early fall (fig. 10).
Figure 10. Seasonal flood-climate regions in the United States. (Modified from Baldwin and McGuinness, 1963.)
WINTER CONDITIONS IN THE SOUTHWEST
Major regional floods such as those of 1993 are generally associated with repeated frontal passages in winter (House and Hirschboeck, 1997). These regional storms may cause the greatest loss because they effect several drainage basins and the severity of flooding increases rather than dissipates downstream. Ely and others (1993) studied weather patterns during modern floods (fig. 6) and concluded that "During winter floods, the typical circulation patterns features a strong southerly displaced low-pressure anomaly off the California coast and, to the north a high pressure anomaly over the Aleutians of the Gulf of Alaska". House and Hirscboeck (1997) point out that the steering of alternatively warm and cold storms into the southwest along a southerly displaced storm track is "enhanced by an active subtropical jet stream, common during El Niño-Southern oscillation periods".
MIXED POPULATIONS OF FLOODS
The previous sections of this article have illustrated the variety of climatic processes that cause floods in the United States. An important reason for understanding these processes is that a better knowledge of the processes may improve predictions of future flooding.
Floods caused by distinctly different climatic processes commonly have distinctly different magnitude and frequency relations. A variety of studies have examined the nature of these differences by separating flood data for a station into two or more populations on the basis of the climatic causes of the floods (U.S. Army Corps of Engineers, 1958; Elliott and others, 1982; Jarrett and Costa, 1982; Waylen and Woo, 1982; Hirschboeck, 1987b). Results of these studies for different regions have indicated that floods caused only by snowmelt, by rain on snow, and only by rain form distinct populations; floods caused by rain on snow or only by rain tend to have larger magnitudes than do floods caused only by snowmelt. In parts of the arid Southwest, floods caused by precipitation from frontal passages in the winter tend to be larger than floods caused by precipitation from convectional storms in the summer. In the Southwest and Northeast, floods caused by precipitation from tropical cyclones tend to have a greater magnitude than do floods caused by precipitation from storms other than tropical cyclones. Floods caused by precipitation from tropical cyclones commonly include the peak flow of record. By defining regions where climatically separated flood populations are distinct, and by effectively applying appropriate statistical procedures to compute flood probabilities from mixed populations, hydrologists generally can determine more accurate estimates of the 10-, 50-, or 100-year floods.
Ely and others (1993) also documented and determined the ages of 251 floods in 19 river basins in the Southwestern United States (fig. 5). Figure 5 shows that intervals of flooding are correlated with periods of cool, moist climate and frequent El Niño events. In regard to possible global warming, they also note that "transitional climate events may contain the highest hydrologic variance" (possible intervals with drought and floods). Curiously from 3600-2200 years B.P., no large floods were documented, and a sharp drop in the number of large floods occurred between 600 and 800 years B.P. (fig 5).
The meteorological processes that produce flood-causing precipitation and runoff in the United States occur within a much larger climatic context that is global in extent. The seasonal and geographic distributions of flood-causing precipitation and runoff are related to the large-scale, general circulation of the atmosphere, which determines the seasonal availability and large-scale delivery pathways of atmospheric moisture. Within a given season, the frequency, typical locations, and degree of persistence of the meteorological processes that release the delivered moisture to cause a flood are influenced by large-scale, atmospheric circulation patterns that develop over areas that are much larger than the flood-affected region. Furthermore, the large-scale climatic framework that influences the occurrence of floods can have a continuity that is much longer than the period of flooding. This occurs when the climate-related, land-surface conditions that affect flood runoff--such as saturated soils or extensive snow cover--develop over a period of several weeks or months.
The role of climate in the occurrence of floods varies from region to region in the United States. Some regions are dominated by several different sources of flood-causing precipitation during the year, whereas other regions are dominated by only one or two. Floods caused by distinctly different climatic processes have distinctly different magnitude and frequency relations. Continued investigation into the effects of multiple sources of flood-producing precipitation will provide hydrologists with more accurate flood estimates, in addition to a better understanding of the flooding process itself.
For the dry Southwestern U.S. specifically, floods can wreck havoc. Their severity and frequency appears to co-vary with El Niño. Studies of El Niño-Southern Oscillation through time suggests changes in the 1990šs that may be related to global warming (Trenberth and Hoar, 1996). If global warming occurs and El Niño increases with global warming, then winter flooding might be expected to increase. Studies such as that by Ely and others (1993) of past frequency of large floods and their association with changes in winter or summer climate can better assess what flood frequencies may be associated with different patterns of climate change.
Aldridge, B.N., and Eychaner, J.H., 1982, Floods of October 1977 in southern Arizona and March 1978 in central Arizona: U.S. Geological Survey Open-File Report 82-687, 167 p.
Augustine, J.A., and Howard, K.W., 1988, Mesoscale convectivc complexes over the United States during 1985: Monthly Weather Review, v. 116, no. 3, p. 685-701.
Bailey, J.F., Patterson, J.L., and Paulhus, J.L., 1975, Hurricane Agnes rainfall and floods, June-July 1972: U.S. Geological Survey Professional Paper 924, 403 p.
Baldwin, H.L., and McGuinness, C.L., 1963, A primer on ground water: U.S. Geological Survey Special Book, p. 21.
Bryson, R.A., and Hare, F.K., 1974, The climates of North America, in Bryson, R.A., and Hare, F.K., eds., Climates of North America, world survey of climatology, v. 11: Amsterdam, The Netherlands, Elsevier, p. 1-47.
Burmeister, I.L., 1985, June 1984 floods on the Missouri River and tributaries, in U.S. Geological Survey, National water summary 1984--Hydrologic events, selected water-quality trends, and ground-water resources: U.S. Geological Survey Water-Supply Paper 2275, p. 40-41.
Chappell, C.F., 1986, Quasi-stationary convective events, in Ray, P.S., ed., Mesoscale meteorology and forecasting: Boston, Mass., American Meteorological Society, p. 289-310.
Chin, E.H., Skelton, John, and Guy, H.P., 1975, The 1973 Mississippi River basin flood--Compilation and analyses of meteorologic, streamflow, and sediment data: U.S. Geological Survey Professional Paper 937, 137 p.
Church, Michael, 1988, Floods in cold climates, in Baker, V.R., Kochel, R.C., and Patton, P.C., eds., Flood geomorphology: New York, John Wiley, p. 205-229.
Clark, G.M., Jacobson, R.B., Kite, J.S., and Linton, R.C., 1987, Storm-induced catastrophic flooding in Virginia and West Virginia, November, 1985, in Mayer, Larry, and Nash, David, eds., Catastrophic flooding: Boston, Mass., Allen and Unwin, p. 355-379.
Costa, J.E., 1987, Hydraulics and basin morphometry of the largest flash floods in the conterminous United States: Journal of Hydrology, v. 93, no. 3/4, p. 313-338.
Court, Arnold, 1980, Tropical cyclone effects on California: National Oceanic and Atmospheric Administration Technical Memorandum NWS WR-159, 41 p.
Court, Arnold, and Griffiths, J.F., 1983, Thunderstorm climatology, in Kessler, Edwin, ed., Thunderstorm morphology and dynamics: Norman, University of Oklahoma Press, p. 9-39.
Cry, G.W., 1967, Effects of tropical eyclone rainfall on the distribution of precipitation over the eastern and southern United States: U.S. Department of Commerce, Environmental Science Services Administration Professional Paper 1, 67 p.
Cry, G.W., Caso, E.L., and Jarvinen, B.R.., 1981, Tropical cyclones of the North Atlantic Ocean 1871-1980 (revised): Asheville, N.C., National Oceanic and Atmospheric Administration, National Climatic Center, 174 p.
Douglas, A.V., Blackmon, R.H., and Englehart, P.J., 1987, Mesoscale convective complexes in extreme western Mexico--A regional response to broadscale drought, in Climate Diagnostics Workshop, 11th, Champaign, 111., 1987, Proeeedings: National Oceanic and Atmospherie Administration, p. 129-140.
Eagleman, J.R., 1983, Severe and unusual weather: New York, Van Nostrand Reinhold, 372 p.
Edelen, G.W., Jr., and Miller, J.F., 1976, Floods of March-April 1973 in southeastern United States: U.S. Geological Survey Professional Paper 998, 283 p.
Edelen, G.W., Jr., Wilson, K.V., Hawkins, J.R., Miller, J.F., and Chin, E.H., 1986, Floods of April 1979, Mississippi, Alabama, and Georgia: U.S. Geological Survey Professional Paper 1319, 212 p.
Elliot, J.G., Jarrett, R.D., and Ebling, J.L., 1982, Annual snowmelt and rainfall peak-flow data on selected foothills region streams, South Platte River, Arkansas River, and Colorado River basins, Colorado: U.S. Geological Survey Open-File Report 82-426, 88 p.
Ely, L.L., Enzel, Yehoda, Baker, V.R., and Cayan, D.R., 1993, A 5,000-year record of extreme floods and climate change in the southwestern United States: Science, v. 262, p. 410-412.
Fontaine, R.A., 1987, Flood of April 1987 in Maine, Massachusetts, and New Hampshire: U.S. Geologieal Survey Open-File Report 87-460, 35 p.
Glancy, P.A., and Harmsen, Lynn, 1975, A hydrologic assessment of the September 14, 1974, flood in Eldorado Canyon, Nevada: U.S. Geological Survey Professional Paper 930, 28 p.
Glatfelter, D.R., Butch, G.K., and Stewart, J.A., 1984, Floods of March 1982, Indiana, Michigan, and Ohio: U.S. Geological Survey Water-Resources Investigations 83-4201, 40 p.
Gray, W.M., 1979, Hurricanes--Their formation, structure, and likely role in the tropical circulation, in Shaw, D.B., ed., Meteorology over the tropical oceans: London, Royal Meteorological Society, p. 155-218.
Hansen, E.M., and Schwarz, F.K., 1981, Meteorology of important rainstorms in the Colorado River and Great Basin drainages: National Oceanic and Atmospheric Administration Hydrometeorological Report 50, 167 p.
Hansen, E.M., Fenn, D.D., Schreiner, L.C., Stodt, R.W., and Miller, J.F., 1988, Probable maximum precipitation estimates--United States between the Continental Divide and the 103rd meridian: National Oceanic and Atmospheric Administration Hydrometeorological Report 55A, 242 p.
Harris, D.D., and Nakahara, R.H., 1980, Flood of February 20, 1979, in Hilo, Kau, and Puna Districts, Island of Hawaii: U.S. Geological Survey Circular 81, 28 p.
Hauth, L.D., 1985, Floods in central, southwest Oklahoma, October 17-23, 1983: U.S. Geological Survey Open-File Report 85-494, 17 p.
Hauth, L.D., Carswell, W.J., Jr., and Chin, E.H., 1981, Floods in Kansas City, Missouri and Kansas, September 12-13, 1977: U.S. Geological Survey Professional Paper 1169, 47 p.
Hayden, B.P., 1988, Flood climates, in Baker, V.R., Kochel, R.C., and Patton, P.C., eds., Flood geomorphology: New York, John Wiley, p. 13-26.
Hirschboeck, K.K. 1991, Climate and floods, in National Water Summary 1988-89, Hydrologic Events of Floods and Droughts: U.S. Geological Survey Water Supply paper 2375, p.67-88.
____1987a, Catastrophic flooding and atmospheric circulation anomalies, in Mayer, Larry, and Nash, David, eds., Catastrophic flooding: Boston, Mass., Allen and Unwin, p. 23-56.
____1987b, Hydroclimatically defined mixed distributions in partial duration flood series, in Singh, V.P., ed., Hydrologic frequency modeling: Boston, Mass., D. Reidel, p. 192-205.
House, P.K., and Hirschboeck, K.K. 1997 (in press). Hydroloclimatological and peleohydrological context of extreme winter flooding in Arizona, 1993, in Larson, R.A., and Slosson, J.E., eds., Effects of the 1992-1993 Winter Storms on Southern California and Arizona: Boulder, Colorado, Geological Society of America Engineering Geology Case Histories Number 12, p.
Hoxit L.R., Maddox, R.A., Chappell, C.F., and Brua, S.A., 1982, Johnstown--Western Pennsylvania storm and floods of July 19-20, 1977: U.S. Geological Survey Professional Paper 1211, 68 p.
Jarrett, R.D., and Costa, J.E., 1982, Multidisciplinary approach to the flood hydrology of foothills streams in Colorado, in Johnson, A.J., and Clark, R.A., eds., International Symposium on Hydrometeorology, Denver, Colo., 1982, Proceedings: American Water Resources Association, p. 565-569.
Lamke, R.D., and Bigelow, B.B., 1987, Floods of October 1986 in south-central Alaska: U.S. Geological Survey Open-File Report 87-391, 31 p.
Lowham, H.W., and Druse, S.A., 1986, Storm and flood of August 1, 1985, in Cheyenne, Wyoming, in U.S. Geological Survey, National water summary 1985- Hydrologic events and surface-water resources: U.S. Geological Survey Water-Supply Paper 2300, p. 41-42.
Lumia, Richard, Burke, P.M., and Johnston, W.H., 1987, Flooding of December 29, 1984, through January 2, 1985, in northern New York State, with flood profiles of the Black and Salmon Rivers: U.S. Geological Survey Water-Resources Investigations Report 86-4191, 53 p.
Maddox, R.A., 1980, Mesoscale convective complexes: American Meteorological Society Bulletin, v. 61, no. 11, p. 1374-87.
____1983, Large-scale meteorological conditions associated with midlatitude, mesoscale convective complexes: Monthly Weather Review, v. 111, no. 7, p. 1475-1493.
Maddox, R.A., Canova, Faye, and Hoxit, L.R., 1980, Meteorological characteristics of flash floods over the western United States: Monthly Weather Review, v. 108, no. 11, p. 1866-1877.
Maddox, R.A., Chappell, C.F., and Hoxit, L.R., 1979, Synoptic and mesoscale aspects of flash flood events: American Meteorological Society Bulletin, v. 60, no. 2, p. 115-123.
Maddox, R.A., Howard, K.W., Bartels, D.L., and Rodgers, D.M., 1986, Mesoscale convective complexes in the middle latitudes, in Ray, P.S., ed., Mesoscale meteorology and forecasting: Boston, American Meteorological Society, p. 390-414.
Maddox, R.A., Rodgers, D.M., and Howard, K.W., 1982, Mesoscale convective complexes over the United States during 1981--Annual summary: Monthly Weather Review, v. 110, no. 10, p. 1501-1514.
Mather, J.R., ed., 1964, Average climatic water balance data of the continents, pt. VII, United States of Publications in climatology: Centerton, N.J., C.W. Thornthwaite Associates Laboratory of Climatology, v. XVII, no. 3, p. 419-615.
McCain, J.F., Hoxit, L.R., Maddox, R.A., Chappell, C.F., and Carcena, Fernando, 1979, Storm and flood of July 31-August 1, 1976, in the Big Thompson River and Cache la Poudre River basins, Larimer and Weld Counties, Colorado--Meteorology and hydrology in Big Thompson River and Cache la Poudre River basins: U.S. Geological Survey Professional Paper 1115-A, p. 1-85.
Meier, M.F., 1986, Snow, ice, and climate--Their contribution to water supply, in U.S. Geological Survey, National water sumrnary 1985--Hydrologic events and surface-water resources: U.S. Geological Survey Water-Supply Paper 2300, p. 69-82.
Morrill, R.A., Chin, E.H., and Richardson, W.S., 1979, Maine coastal storm and flood of February 2, 1976: U.S. Geological Survey Professional Paper 1087, 20 p.
Muller, R.A., and Faiers, G.E., eds., 1984, A clirnatic perspective of Louisiana floods during 1982-1983: Baton Rouge, Louisiana State University, Geoscience Publications, 48 p.
Parrett, Charles, Carlson, D.D., and Craig, G.S., 1984, Floods of May 1978 in southeastern Montana and northeastern Wyoming: U.S. Geological Survey Professional Paper 1244, 74 p.
Rabenhorst, T.D., 1981, An atlas of the United States: University of Maryland Baltimore County, Department of Geography, 83 p.
Ray, P.S., ed., 1986, Mesoscale meteorology and forecasting: Boston, American Meteorological Society, 793 p.
Reed, R.J., 1960, Principal frontal zones of the Northern Hemisphere in winter and summer: American Meteorological Society Bulletin, v. 41, no. 11, p. 591-598.
Reitan, C.H., 1960, Distribution of precipitable water vapor over the continental United States: American Meteorological Society Bulletin, v. 41, no. 2, p. 79-87.
____1974, Frequencies of cyclones and cyclogenesis for North America, 1951-1970: Monthly Weather Review, v. 102, no. 12, p. 861-868.
Rodgers, D.M., and Howard, K.W., 1983, Mesoscale convective complexes over the United States during 1982--Annual summary: Monthly Weather Review, v. 111, no. 12, p. 2363-2369.
Runner, G.S., and Chin, E.H., 1980, Flood of April 1977 in the Appalachian region of Kentucky, Tennessee, Virginia, and West Virginia: U.S. Geological Survey Professional Paper 1098, 43 p.
Saarinen, T.F., Baker, V.R., Durrenberger, Robert, and Maddock, Thomas, Jr., 1984, The Tucson, Arizona, flood of October 1983: Washington, D.C., National Academy Press, National Research Council, Committee on Natural Disasters, 112 p.
Sauer, V.B., and Fulford, J.M., 1983, Floods of December 1982 and January 1983 in central and southern Mississippi River basin: U.S. Geological Survey Open-File Report 83-213, 41 p.
Schroeder, E.E., Massey, B.C., and Chin, E.H., 1987, Floods in central Texas, August 1-4, 1978: U.S. Geological Survey Professional Paper 1332, 39 p.
Schwarz, F.K., Hughes, L.A., Hansen, E.M., Petersen, M.S., and Kelly, D.B., 1975, The Black Hills-Rapid City flood of June 9-10, 1972--A description of the storm and flood: U.S. Geological Survey Professional Paper 877, 47 p.
Smith, W.S., 1986, The effects of eastern North Pacific tropical cyclones on the southwestern United States: National Oceanic and Atmospheric Administration Technical Memorandum NWS WR-197, 229 p.
Smith, W.S., and Gall, R.L., 1989, Tropical squall lines of the Arizona monsoon: Monthly Weather Review, v. 117, no. 7, p. 1553-1569.
Thomas, C.A., and Lamke, R.D., 1962, Floods of February 1962 in southern Idaho and northeastern Nevada: U.S. Geological Survey Circular 467, 30 p.
Trenberth, K.E., and Hoar, T.J., 1996, The 1990-1995 El Niño-Southern Oscillation event longest on record: Geophysical Research Letters, v. 23, No. 1, p. 57-60.
U.S. Army Corps of Engineers, 1958, Frequency of New England floods: Sacramento District, Research Note 1, 4 p.
____1966, Hurricane Betsy, 8-11 September 1965-- After-action report: New Orleans District, 73 p.
____1970, Report on Hurricane Camille, 14-22 August 1969: New Orleans District, 143 p.
U.S. Weather Bureau, 1964, Frequency of maximum water equivalent of March snow cover in north central United States: U.S. Weather Bureau Technical Paper 50, 24 p.
Waananen, A.O., Harris, D.D., and Williams, R.C., 1971, Floods of December 1964 and January 1965 in the Far Western States: U.S. Geological Survey Water-Supply Paper 1866-A, 265 p.
Waylen, Peler, and Woo, Ming-Ko, 1982, Prediction of annual floods generated by mixed processes: Water Resources Research, v. 18, no. 4, p. 1283-1286.
Webber, E.E., 1982, Flood of June 13-15, 1981, in the Blanchard River basin, northwestern Ohio: U.S. Geological Survey Water-Resources Investigations Report 82-4044, 32 p.
Wendland, W.M., and Bryson, R.A., 1981, Northern hemisphere airstream regions: Monthly Weather Review, v. 109, no. 2, p. 255-270
Wilson, K.V., and Hudson, J.W., 1969, Hurricane Camille tidal floods of August 1969 along the Gulf Coast, Kiln quadrangle, Mississippi: U.S. Geological Survey Hydrologic Investigations Atlas HA-397.