Floods

Floods are extreme conditions of flowing water. They generally occur because of inordinate amounts of rainfall or snowmelt, but floods may also result from other causes, such as dam failures or volcanic eruptions. Floods play a major role in shaping river systems, and their occurrence is critical to the human use of riparian lands.

Causes of Floods

Floods involve extremely large flows of water in rivers and streams. Some technical hydrological definitions of floods refer to stages, or heights, of water above some reference level, such as the banks of a river channel. In practice, however, floods can be thought of as any extreme flow of water that exceeds usual experience, often damaging or threatening life and property.

Floods may be caused by a variety of physical factors, including dam failures and the subsidence of land. The most common kind of flood occurs when excessive precipitation and physical factors of the land combine to produce maximum runoff of water. Precipitation can yield runoff directly, or water from melting snow can produce the flow. Very intense flash floods are usually associated with heavy, short-duration rainfall from thunderstorms. Such rainfall easily overwhelms the infiltration capacity of the ground, and water rapidly runs off from the local drainage area or watershed into adjacent stream channels. If enough water concentrates in the channel to exceed the carrying capacity of the channel, it will constitute a flood.

Physical factors on the land surface also determine the rate of concentration of floodwater in stream channels. For example, less permeable soils will allow water to run off faster, as will a lack of vegetation. Artificial enhancement of runoff occurs when slopes are covered by impermeable materials. This situation commonly occurs in construction when the natural ground surface is replaced by buildings and pavement, resulting in extreme enhancement of runoff. In cities, the yield of floodwater from paved surfaces may be ten times greater than the same storm’s yield under natural conditions. In this way, human construction tends to exacerbate flooding problems in urban areas.

Sediment and Bedrock

As the flow of water increases in a stream channel, it is usually associated with considerable sediment. If such waterborne sediment also composes extensive deposits adjacent to the stream channel, the river is termed alluvial. An alluvial river commonly has a floodplain of deposited sediment adjacent to the channel. The floodplain is really an intimate part of the river system, as sediment is added to it every time the river rises above its banks. As the water rises, the river’s depth rapidly increases, causing an increased ability to transport sediment that is eroded from the bed and banks. If the stream is appropriately loaded with sediment, it will be deposited when the banks are overtopped, and the width of the flow greatly increases. If not enough sediment is supplied by erosion, however, the increasingly energetic flood flows will be erosive, attacking the banks, widening the channel, and restoring the appropriate sediment load to the stream.

When the bed and banks of a river are composed of bedrock, these adjustments of sediment load to flow energy cannot occur. Because the bedrock can be very resistant to erosion, the energy level in the flow can rise spectacularly without being damped by sediment. The excess energy of such sediment-impoverished floods goes into the development of turbulence. Turbulence at high energy levels takes on an organized structure of powerful vortices that produce immense pressure changes. These pressure effects may be sufficient to erode the bedrock boundary by a “plucking” action.

A famous geological controversy once surrounded the problem of bedrock erosion by great floods. In the 1920s, University of Chicago geologist J. Harlen Bretz proposed that immense tracts of eroded basalt bedrock in Washington State had been created by a catastrophic glacial flood. Bretz subsequently showed that the fascinating landforms of that region, known as the Channeled Scablands, were created when a great lake impounded by glacial ice burst. The lake, glacial Lake Missoula, had been more than six hundred meters deep at its ice dam. It took Bretz nearly fifty years to convince his many critics that catastrophic flooding could explain all the bizarre features of the Channeled Scablands. Geologists came to realize that Bretz's observations were completely consistent with the physics of catastrophic flooding is completely consistent with Bretz’s observations. Missoula flood flows moved at depths of one hundred to two hundred meters and velocities of twenty to thirty meters per second. The power (rate of energy expenditure) per unit area for such flows is thirty thousand times that for a normal river, such as the Mississippi, in flood. The reason for such immense flow power is that the Missoula flooding occurred on very steep slopes, giving the water great potential energy when the dam burst.

The Missoula floods occurred during the last ice age, more than twelve thousand years ago. There are, however, modern examples of glacial floods called jökulhlaups. In Iceland, jökulhlaups occur where glaciers overlie active volcanoes. Volcanic heat releases water that is stored in subsurface reservoirs by melting the overlying ice. Lakes may also form adjacent to the ice masses. Because ice is less dense than water, such juxtapositions are inherently unstable. When the pressure is high enough, the water may lift the ice dam and burst out from beneath the glacier. The jökulhlaups move house-sized boulders and transport immense quantities of sediment.

Discharge and Recurrence Interval

The volume of water released by a flood per unit of time is termed its discharge. This quantity is the magnitude of the flow that will potentially inundate an area. The chances of experiencing floods of different magnitudes are expressed in terms of frequency. Large, catastrophic floods have a low frequency, or probability of occurrence; smaller floods occur more often. The probability of occurrence for a flood of a given magnitude can be expressed as the odds, or percent chance, of the recurrence of one or more similar or bigger floods in a certain number of years. Analyses of flood magnitude and frequency are achieved by measuring floods and statistically analyzing the data. Results are expressed in terms of the probability of a given discharge being equaled or exceeded in any one year. The reciprocal of this probability is the return period, or recurrence interval, of the flooding, expressed as a number of years.

The concept of a recurrence interval is sometimes confusing; it is simply a statistical measure, based on historical measures of the flood frequency of a river flow, of the likelihood of a flood of a given magnitude occurring over a one-year period. A flood that has a probability of occurrence in one year of 0.1, or 10 percent, is called a ten-year flood because the one-year probability (0.1) multiplied by 10 is equal to 1. By extension, a larger flood that has a probability of occurrence of 0.01, or 1 percent, in a single year is called a hundred-year flood because 0.01 multiplied by 100 is equal to 1. Note that these numbers do not preclude several such floods occurring in a given period; for example, two or more floods of the one-hundred-year magnitude could occur in the same year. Such an event, while unlikely, does have a small probability of occurring. In fact, in August 2017, in advance of the flooding caused by Hurricane Harvey in Houston, Texas—later calculated to be a "thousand-year flood," the first such recorded event in US history, with a 0.001 probability (0.1 percent) of occurring in a single year—meteorologist Eric Holthaus reported that the Houston area had seen four hundred-year floods since May 2015.

Flood magnitude-frequency relationships vary immensely with climatic regions. In the humid-temperate regions of the globe, such as the northern and eastern United States, streamflow is relatively continuous. Even rare floods are not appreciably larger than more common floods. A result of this relationship is that stream channels have developed in size to convey the relatively common, moderate-sized floods with maximum efficiency. In contrast, the more arid regions of the southwestern United States have immensely variable flood responses. Stream channels may be dry most of the time, filling with water only after rare thunderstorms. The flash floods that characterize these streams may also be highly charged with sediment; indeed, the sediment may so dominate the flow that the phenomenon is called mudflow or debris flow rather than streamflow. Because extreme events dominate in these environments, the stream channels have developed in size according to these rare, great floods.

Monsoons and Ancient Floods

In tropical areas, some of the greatest known rainfalls are produced by tropical storms and monsoons. Monsoons are seasonal wet-to-dry weather patterns driven by atmospheric pressure changes over the oceans and continents, which result in alternations between dry periods that inhibit vegetation growth and immensely wet periods that facilitate runoff. Tropical rivers thus have a pronounced seasonal cycle of flooding, and some floods may be immense. These rivers also show channel-size development in accordance with rare, great flows. Some of the most populous places on Earth are situated on seasonal tropical rivers, such as the Ganges and Brahmaputra Rivers in India and Bangladesh. Immense tragedies have occurred when a particularly severe monsoon or tropical storm has produced especially great floods. In 1974, monsoon-related flooding in Bangladesh killed 2,500 people. This pales in comparison to a flood in 1970 that took 500,000 lives and another in 1991 that killed more than 100,000 people. India's National Disaster Management Authority has estimated that an average of 1,600 lives are lost annually due to flooding, with the greatest loss of life—11,316 people—having occurred in 1977.

Studies of ancient floods (paleofloods) through geological reconstruction of past discharges show that monsoons and other flood-generating systems have varied in the past. Between about ten thousand and five thousand years ago, floods were apparently much more intense in many world areas on the boundaries between the tropics and the mid-latitude deserts. These intense floods may have been related to long-term glacial-to-interglacial cycles. Because of modern increases in atmospheric carbon dioxide and other greenhouse gases, it appears quite likely that tropical floods may again become more intense if this process has not already begun. Such a situation could have grave consequences for flood-prone tropical countries.

Study of Floods

Hydrologists study floods by measuring flow in streams. Measurements are taken at stream gauges, locations at which mechanical devices are used to record the water level, or stage, of the river. To transform these stage measurements into discharge values, the hydrologist must perform a rating of the stream gauge. This is accomplished by measuring velocities in the stream channel during various flow events. A velocity meter with a rotor blade calibrated to the flow rate is used for this purpose. When the average measured velocity of the stream channel is multiplied by the cross-sectional area of the channel, the result is the discharge for that flow event. When several flows at different stages are measured, the data are used to generate a rating curve for the gauge. This curve shows discharges corresponding to any stage measured at the gauge.

The discharge values obtained at a stream gauge are collected over many years, constituting a record that can then be used in flood-frequency analysis. Several statistical procedures can be employed to plot the flood experience and to extrapolate to ideal values of the ten-year flood, the one-hundred-year flood, and so on. The discharges associated with these recurrence intervals are then used as design values for hazard assessment, dam construction, and other flood controls. “Flood control” is probably a misnomer, however, because flows larger than the design floods are always possible. Flood control really involves providing various degrees of flood protection.

Another approach to evaluating extreme flood magnitudes involves careful study of the precipitation values that generate the greatest known floods. By transposing the patterns of known extreme storms to other areas, scientists can, in theory, calculate what the runoff would be from hypothetical great storms. Such calculations involve the use of a rainfall-runoff model. These models are prevalent in hydrology because they can be easily programmed. The models give idealized predictions of how water from a given storm would concentrate. The flood discharge modeled from the assumed maximum rainfall is called the probable maximum flood.

Unfortunately, there are problems inherent to both of these traditional hydrological approaches to flood studies. Both make assumptions in calculating potential flood flows. Another procedure is to study the natural records of ancient floods, or paleofloods, that are preserved in geological deposits or in erosional features on the land. It has been found for some sections of bedrock that nonalluvial rivers act as natural recorders of extremely large flood events. These natural flood gauges can be interpreted only by detailed studies that combine geological analysis with hydraulic calculations of the ancient discharges.

Paleoflood hydrology generates real data on the largest floods to occur in various drainages over several millennia. The data include the floods’ ages, or the periods in which they occurred, and their discharges. The information can be used directly in a flood-frequency analysis, or it can be used to assess the probable validity of extrapolations from conventional data on smaller floods. Paleoflood data can also be compared with probable maximum flood estimates. In this way, the expense of overdesign and the danger of underdesign can be avoided in flood-related engineering projects.

Significance

People have lived with floods since the beginning of civilization. The first great civilizations on Earth developed along the fertile but flood-prone valleys of rivers such as the Nile, the Tigris and the Euphrates, the Indus, and the Yangtze and the Huang He. Various means of coping with floods have been documented since the biblical accounts of Noah. Early societies merely avoided zones that tradition told them were hazardous. It has only been in the modern era that large cities have systematically developed on immense tracts of flood-prone lands. Thus the natural process of flooding has become an unnatural hazard to humans.

Floods can have immense consequences for both human life and infrastructure. In the 1930s, a national US program began to respond to such problems by constructing large dams. Despite, or perhaps because of, this expensive effort, flood damage to life and property is much greater today than it was in the 1930s. A single tropical storm system, Hurricane Agnes, killed 128 people overall and generated nearly $3 billion in flood damage to the northeastern United States in 1972; in 2011, the flooding of the Ohio, Mississippi, and Missouri Rivers in April and May, plus the flooding caused by Hurricane Irene and Tropical Storm Lee in August and September, caused a total of 108 deaths and more than $8 billion in damages in the United States; and New Orleans may never fully recover from the flooding caused by Hurricane Katrina in 2005, which resulted in as many as 1,464 deaths and $70 billion in damages in that city alone. The flooding caused by Hurricane Harvey in 2017 was estimated to have caused between $65 billion and $190 billion worth of damage in Texas, the upper range of which would make it the costliest natural disaster in US history, while Hurricane Irma, which followed just days after, inflicted an additional $50 billion to $100 billion in damages in the southeastern United States. Then, in 2021, Hurricane Ida made landfall in Louisiana and became the second-worst hurricane to hit the state since Katrina. The storm's direct impact and subsequent flooding, which also affected the larger US as well as the Caribbean and South America, led to estimates of over $70 billion in damage within the US.

Yellowstone National Park and regions of Montana experienced historical flooding in 2022 after record levels of rainfall overwhelmed the Yellowstone River. The resulting floods washed away the landscape, roads, bridges, and homes. The high-water mark of the 2022 Yellowstone flood, which damaged hundreds of homes and caused the evacuation of over 10,000 park visitors, was higher than any other on record.

In 2024, Tropical Storm Helene inundated western North Carolina, eastern Tennessee, South Carolina, and other parts of the Southeastern US with historically high rainfall after making landfall as a Category 4 hurricane in Florida. The storm caused catastrophic flooding in the region, in which rivers flooded their banks, hundreds of roads were washed out, whole towns were destroyed, and more than one hundred people died.

Elsewhere, flooding is a particularly serious issue in South Asia, where monsoon season is an annual occurrence, and flood-control measures are largely limited to Dutch-style embankments that do not take into account the greater sediment present in the major rivers. In 2007, unusually heavy rains during the monsoon season caused thirty-seven major floods throughout several South Asian countries, including India, Bangladesh, Pakistan, and Nepal, killing more than 4,000 people and displacing or otherwise affecting more than 45 million. In the summer of 2017, flooding in India, Bangladesh, and Nepal caused more than 1,200 deaths and destroyed close to a million houses, affecting more than 41 million people.

Extreme flooding continued into the next decade. In July 2021, record rainfall caused flooding in China's Henan Province that led to the deaths of nearly 400 people and the evacuation of 800,000 others. Later that same month, flooding from heavy rain in Maharashtra, India, led to over 250 deaths and the evacuation of over 375,000 residents. One year later, floods in KwaZulu-Natal, South Africa, following heavy rains, led to over 450 deaths and an estimated $1.5 billion in damages. The nature of the floods, each caused by extreme levels of rainfall, led to renewed discussion on the link between climate change and worsening global weather patterns.

In addition to the dangers posed by flooding itself, the aftermath of a major flood can pose serious health risks to residents. Remaining floodwater may be contaminated with sewage and chemicals, and that contamination can be transferred to flood-damaged homes. Mold, which grows quickly in warm, damp conditions, can cause or trigger respiratory issues. Standing water attracts mosquitoes, increasing the risk of mosquito-borne illnesses. In addition, the stress, anxiety, and depression caused by the event can lead to serious mental health issues for survivors.

One trend in flood management has been to manage flood-hazard zones with multiple approaches that respond to the nature of the flood risk. The river is treated as a whole integrated system, rather than as individual segments, for engineering design. Management alternatives for this system are not limited solely to structural controls, such as dams and levees. Instead, options are considered for land-use adjustment. Flood-prone lands can be used for parks, greenbelts, and bikeways instead of industrial warehouses, stores, and housing. Even when construction must be done on floodplains, it may be possible to make provision for flood risks. Warehouses, for example, can be organized for the rapid transfer of materials to second stories or to temporary, safe storage sites. Such adjustments require accurate and timely warning systems that involve measuring rainfall in headwater areas and rapidly predicting the flood consequences to downstream sites at risk.

There is a general need to educate the public that floodplains are a natural part of rivers. Living on a floodplain is really choosing to play a game of “floodplain roulette,” in that it is known with certainty that a flood will happen, but it is not known when that flood will happen. For this reason, the most accurate and reliable methods of evaluating flood magnitudes and frequencies are necessary. The choices made for land use on floodplains cannot be based merely on idealized theories of how floods behave.

Principal Terms

discharge: the volume of water moving through a given flow cross-section in a given unit of time

flash floods: rises in water level that occur unusually rapidly, generally because of especially intense rainfall

flood: a rising body of water that overtops its usual confines and inundates land not usually covered by water

hydrology: the branch of science dealing with water and its movement in the environment

jökulhlaup: a flood produced by the release of water sequestered by a glacier, most often due to the failure of some type of glacial dam or to subglacial volcanic activity

monsoon: a seasonal, reversing pattern of wind between warm ocean bodies and landmasses

recurrence interval: the average time interval in years between occurrences of a flood of a given magnitude in a measured series of floods

runoff: that part of precipitation that flows across the land and eventually gathers in surface streams

Bibliography

Aldrete, Gregory S. Floods of the Tiber in Ancient Rome. Johns Hopkins UP, 2007.

American Society of Civil Engineers, Hurricane Katrina External Review Panel. The New Orleans Hurricane Protection System: What Went Wrong and Why. ASCE, 2007. ASCE Library, doi:10.1061/9780784408933. Accessed 6 Sept. 2017.

Baker, Victor R., editor. Catastrophic Flooding: The Origin of the Channeled Scabland. Dowden, Hutchinson & Ross, 1981.

Baker, Victor R., et al., editors. Flood Geomorphology. John Wiley & Sons, 1988.

Christensen, Jen. "The Hidden Health Dangers of Flooding." CNN, 31 Aug. 2017, www.cnn.com/2017/08/27/health/health-consequences-flood-waters/index.html. Accessed 22 Sept. 2017.

Comerio, Mary C. Disaster Hits Home: New Policy for Urban Housing Recovery. U of California P, 1998.

Dasgupta, KumKum. "Why India Is Failing to Minimise Monsoon Flood Destruction." Hindustan Times, 26 July 2017, www.hindustantimes.com/analysis/why-india-is-failing-to-minimise-monsoon-flood-destruction/story-4qC6DuVWacsebWBwnvLbjI.html. Accessed 22 Sept. 2017.

Freitag, Bob, et al. Floodplain Management: A New Approach for a New Era. Island Press, 2009.

Hassan, Adeel. "What We Know About Hurricane Helene's Destruction So Far." The New York Times, 29 Sept. 2024, www.nytimes.com/2024/09/29/us/helene-destruction-florida-north-carolina.html. Accessed 30 Sept. 2024.

Holthaus, Eric. "A Texas-Size Flood Threatens the Gulf Coast, and We're So Not Ready." Grist, 23 Aug. 2017, grist.org/article/a-texas-size-flood-threatens-the-gulf-coast-and-were-so-not-ready/. Accessed 22 Sept. 2017.

Irfan, Umair. "The Stunning Price Tags for Hurricanes Harvey and Irma, Explained." Vox, 18 Sept. 2017, www.vox.com/explainers/2017/9/18/16314440/disasters-are-getting-more-expensive-harvey-irma-insurance-climate. Accessed 22 Sept. 2017.

Knighton, David. Fluvial Forms and Processes: A New Perspective. Rev. and updated ed., Hodder Arnold, 1998.

Lui, Kevin. "Severe Flooding in South Asia Has Caused More Than 1,200 Deaths This Summer." Time, 29 Aug. 2017, time.com/4921340/south-asia-floods-india-mumbai-bangladesh-nepal/. Accessed 22 Sept. 2017.

O’Loughlin, Karen Fay, and James F. Lander. Caribbean Tsunamis: A 500-Year History from 1498–1998. Kluwer Academic Publishers, 2003.

Orsi, Jared. Hazardous Metropolis: Flooding and Urban Ecology in Los Angeles. U of California P, 2004.

Philippi, Nancy S. Floodplain Management: Ecologic and Economic Perspectives. R. G. Landes, 1996.

Prothero, Donald R. Catastrophes! Earthquakes, Tsunamis, Tornadoes, and Other Earth-Shattering Disasters. Johns Hopkins UP, 2011.

Robbins, Jim, et al. "Flooding Chaos in Yellowstone, a Sign of Crises to Come." The New York Times, 15 June 2022, www.nytimes.com/2022/06/15/us/yellowstone-national-park-floods.html. Accessed 6 July 2022.

Samenow, Jason. "Harvey Is a 1,000-Year Flood Event Unprecedented in Scale." The Washington Post, 31 Aug. 2017, www.washingtonpost.com/news/capital-weather-gang/wp/2017/08/31/harvey-is-a-1000-year-flood-event-unprecedented-in-scale/. Accessed 22 Sept. 2017.

Sampath, Nikita. "Mismatched Flood Control System Compounds Water Woes in Southern Bangladesh." New Security Beat, Environmental Change and Security Program, Woodrow Wilson International Center for Scholars, 3 Jan. 2017, www.newsecuritybeat.org/2017/01/mismatched-flood-control-system-compounds-water-woes-southern-bangladesh/. Accessed 22 Sept. 2017.

Schneider, Bonnie. Extreme Weather: A Guide to Surviving Flash Floods, Tornadoes, Hurricanes, Heat Waves, Snowstorms, Tsunamis and Other Natural Disasters. Palgrave Macmillan, 2012.

Sene, Kevin. Flash Floods: Forecasting and Warning. Springer, 2013.

Sene, Kevin. Flood Warning, Forecasting and Emergency Response. Springer, 2008.

Singh, Vijay P., editor. Hydrology of Disasters. Kluwer Academic Publishers, 1996.

Srivastava, Sanjay, et al. South Asian Disaster Report 2007. SAARC Disaster Management Centre, 2008. SAARC Disaster Management Centre, saarc-sdmc.nic.in/sdr‗p.asp. Accessed 6 Sept. 2017.

Warner, Jeroen Frank, et al., editors. Making Space for the River: Governance Experiences with Multifunctional River Flood Management in the US and Europe. IWA Publishing, 2013.