8.5: Flooding

Why Floods Occur

Floods happen because there is more water flowing in a stream than the normal channel can contain, but it’s not that simple. The most common cause of stream flooding is heavy precipitation, and an example of that is shown on Figure \(\PageIndex{1}\). This is a doppler radar image for an area southwest of Oklahoma City. It shows the degree of radar energy reflected from water droplets in the lower atmosphere. The small pink area may be experiencing rainfall rates of greater than 400 mm/h (which is very heavy rain), while the red areas may be experiencing > 40 mm/h of rain. Streams in that region would have responded with significant discharge rates, and there might have been some localized flooding.

Figure \(\PageIndex{1}\): Doppler Radar Image for an Area southwest of Oklahoma City, OK. The numbers in the legend are in decibels, representing the amount energy returned to the radar station.

Stream discharge rates can also increase dramatically because of the rapid melting of snow and even glacial ice. Rain can contribute to snowmelt, so in many cases flooding results from a combination of these two processes.

The water flowing in a stream and its tributaries comes from two main sources: overland flow (water flowing over the surface of the ground during and following heavy rain or very rapid snow melt), and discharge from groundwater (groundwater at the surface that is added into streams). This second factor only affects the stream if the water table is close to the surface to begin with. In areas with deep water tables, this is not a consideration.

When there is heavy rainfall within a drainage basin some of the water can be expected to flow over the surface, but most of it infiltrates into the ground to become groundwater. This raises the water table, and therefore increases the rate of discharge into the stream. Those two main components contribute to a significant increase in the stream discharge that lasts for several days, and that discharge gradually decreases over time—rapidly at first as the overland flow rate slows, and slowly as the water table subsides and the rate of groundwater discharge into the stream decreases.

Flooding can also happen because of slope failure. If a stream channel is blocked because of slope failure the area upstream may flood, or if a significant amount of water accumulates behind a slope-failure dam, and then the dam fails, the area downstream may flood.

When a stream channel is filled with water to near its capacity, it is said to be at the “bank-full stage”. In the case of streams with wide flood plains the cross-sectional area available for flow will increase dramatically as soon as the stream overtops its bank and starts to flood, and even though there is more water moving than at bank-full stage, the velocity will decrease. As the velocity decreases some of the suspended sediments will be deposited. The coarser material will be deposited close to the normal bank top, forming a natural levee. Finer sediments will be deposited slowly, across the flood plain, over the days in which flood waters remain there, and especially when the flow velocity drops even more.

Figure \(\PageIndex{2}\): A Depiction of a Stream at Bank-Full Stage and at Flood Stage

The natural levees formed by flooding can provide some protection against future flooding, but they can also prevent flood waters from flowing back into the stream, so prolonging the time over which some areas remain flooded.

Flood Recurrence Interval

Providing that we have enough stream discharge data from previous years it is possible to estimate the probability that a flood of a certain size will happen at some time in the future. Flood probability is determined by calculating the flood recurrence interval (Ri), the estimated average time between events of a particular discharge, for any given stream. This type of information is useful for planners that must make decisions about approving proposals for infrastructure projects (buildings, roads etc.) within flood plains, and also for anyone that wants to live near to a river.

Emergency measures organizations use calculations of Ri to report past, current or predicted floods using terms like “100-year flood”. That means that the flood was, or is expected to be, as big as the largest flood in the past 100 years. Another way to think about it is a "100-year flood" occurs on average once every 100 years. This does not mean it happens every 100 years. You could have a 100-year flood two years in a row, and then not again for a long period of time. These floods are larger than "50-year floods" which occur on average, once every 50 years. They are smaller as they occur more often. "10-year floods" are even smaller than that and occur even more often.

Although we can determine a flood probability, we can’t predict when there is going to be a big flood, nor how big it will be, so to minimize damage and casualties we need to be prepared. Some of the ways of doing that are as follows:

• Mapping flood plains and not building within them,
• Building dams where necessary,
• Monitoring the winter snowpack, the weather, and stream discharges,
• Creating emergency plans, and
• Educating the public.

Another important point to remember is that estimation of flood probabilities based on past data relies on the premise that the climate conditions and land-use patterns that produced the historical record are still relevant in the future. As we know, weather can be unpredictable. Furthermore, land use changes may have changed the way water runs off the surface, and that could also change the probability of future large floods.

Examples of Major Flood Events

In late August of 2017, Category 4 Hurricane Harvey made landfall near to Corpus Christi, Texas, and then slowly moved east, remaining almost stationary over the Houston region for several days. In that time more than 1000 mm of rain fell over a large area. There was widespread flooding (Figure \(\PageIndex{3}\)), more than 100 people died, and the amount of damage is estimated to have been $125 billion, mostly in the form of damages to buildings. In terms of the aggregate rainfall amount measured, Harvey was the wettest tropical storm ever to affect the United States.

Figure \(\PageIndex{3}\): Flooding in Pt. Arthur Texas Caused by Hurricane Harvey, August 31st, 2017

The annual number of tropical storms in the Atlantic basin has been increasing in recent decades, mostly because of increased ocean water temperatures as the climate has warmed. The higher water and air temperatures also mean that more water is transferred from the ocean to the atmosphere as a tropical storm evolves, and so we can look forward to a future with more very wet systems like Harvey to cause flooding.

Serious flooding in Canada in June of 2013 was initiated by rapid snow melt in the Rocky Mountains and was worsened by heavy rains due to an anomalous flow of moist air from the Pacific. Rainfall amounts exceeded 200 mm in 36 hours at Canmore, and 325 mm in 48 hours at High River. The discharges of several rivers in the area, reached levels that were 5 to 10 times higher than normal for the late June. Large parts of Calgary and surrounding areas were flooded, and 5 people died (see Figure \(\PageIndex{4}\)). The cost of the 2013 flood is estimated to have been approximately $5 billion, making it the most expensive flood event in Canadian history.

Figure \(\PageIndex{4}\): Flooding in Calgary (June 21, left) and Okotoks (June 20, right) During the 2013 Southern Alberta Flood

The Himalayan region is a common source of snow-melt flooding in China, India, Pakistan, Bangladesh, and other countries in the region, but there have also been many slope-failure related floods in the mountains. In July 2000, a slope-failure in China dammed up the Satluj River, forming a temporary lake. By July 31st the water level in the lake had risen enough to flow over and then quickly erode the dam, releasing a massive flood that is said to have increased the water level of the river by 20 meters. Over 150 lives were lost, 250 houses destroyed, and 20 km of road and 7 bridges washed out. As many as 1000 irrigation, sewerage, flood protection, power installations and water supply systems were damaged.^[1]

Anthropogenic Effects on Flooding

Natural landscapes provide a natural defense against flooding. Plants such as grasses, bushes, and trees keep soil in place which absorbs excess rainfall and allows it to infiltrate into the ground. Mature root systems can also help manage excess water as the roots absorb it. Natural stream systems with developed flood plains and wetlands all serve as places for water to go and be distributed from. Human activities can disrupt these systems, increasing the likelihood and severity of floods.

The development of urban and rural areas hinders these natural systems and increase the frequency and magnitude of flood events. Wetlands are meant to store excess water, but we often drain them or pave over them for building purposes. Water that should be stored there ends up in the stream channels. Levees, meant to keep water in the stream channel and away from the surrounding floodplains, ultimately result in increased flooding downstream. They also result in elevated river levels such that when a levee does fail, massive flooding can result, destroying the structures it was meant to protect. When Hurricane Katrina struck New Orleans in 2005, levee failures flooded the city, killing hundreds of people.

Large agricultural operations can deplete soil of the nutrients which results in reducing its quality. Plants then struggle to survive and die off. Since plants play a key role in protecting areas against flooding, the magnitude and frequency of flood events can increase. Deforestation can remove the large trees which act to keep soil in place. Without them, nutrient rich soil washes away and leaves dry soil in its place which is not suitable for absorbing heavy rainfall. This increases the amount of water that washes into streams, which causes flooding.

Flood Reduction and Control

There are several ways to reduce the risks associated with flooding, including limiting the potential for flooding in the first place, controlling where floodwaters go, and taking steps to limit the amount of damage that floods cause.

One of the best ways to reduce the risk of flooding is to do just the opposite of what we do most of the time in urban areas, namely constructing endless highways, roads and parking lots, and countless buildings. Paved surfaces and roofs do not absorb water effectively, so relatively little of it infiltrates into the ground to become groundwater; instead, most runs off the surface. The runoff coefficients for a range of different surface types are illustrated on Figure \(\PageIndex{5}\). The runoff coefficient is an estimate of the proportion of rain from a significant storm that will flow over a surface to become runoff, rather than infiltrating into the ground.

Figure \(\PageIndex{5}\): Runoff Coefficients for Various Types of Surfaces

Only a small proportion of the rain that falls in parks or forests—5 to 25% of it—becomes surface flow; the rest—75 to 95% of it—infiltrates into the ground and becomes part of the groundwater, and then moves very slowly towards surface drainages, thus delaying and reducing the ultimate size of a flood. For cultivated land, meadows and pasture, 10 to 65% of the precipitation runs across the surface, directly into streams and lakes, while 35 and 90% of the water infiltrates. For roads and streets 70 to 95% flows over the surface, or through ditches and storm sewers directly into surface water bodies, while only 5 to 30% infiltrates. Obviously, increasing the amount of infrastructure (roads, parking lots, buildings) in an area, increases the amount of water that will flow quickly into drainage systems and so will amplify the potential for flooding. On the other hand, conserving forests in their natural state, and creating parks that have minimal infrastructure will reduce the potential for flooding.

Some ways to reduce the flooding potential impact of existing urban regions include reducing the area covered by hard surfaces, making hard surfaces (e.g., roads) more permeable, replacing roads and parking lots with parks and community garden plots, constructing ponds or wetlands to capture runoff, and creating dips and hollows (swales) in landscaped areas to slow the rate of flow and increase the rate of infiltration.

We can control floodwaters by building dams, dykes and channels. The High Aswan Dam in Egypt is a good example of that, although there are significant downsides to the construction of large dams. A dyke or levee is a berm built along one or both sides of a river to prevent flood waters from spilling out onto the floodplain. An artificial channel can be constructed separate from the normal channel to carry a portion of the flow of a river during flood events.

The Mississippi-Missouri-Ohio river system is the largest in North America, with a drainage basin that covers more than 40% of the US and even a small part of Canada. It has great historical significance and has been a critical transportation corridor for over two centuries. The Mississippi is one of the most controlled rivers in the world. There are 64 major dams on the Mississippi, Missouri and Ohio Rivers, and hundreds of others on their tributaries. There are over 5600 km of dykes (a.k.a., levees) along the sides of the rivers, and thousands of other structures that have been built to make the water go where it is wanted, or to keep the shipping channels open. Wing dykes, for example are constructed within the river channel at an angle to the shore and are designed to keep most of the water flowing in the central part of the channel so as to limit the extent to which that part fills with sediment (Figure \(\PageIndex{6}\)). Despite all of the engineering works on the Mississippi system (or in some cases because of them), the tendency for flooding has increased over the past century. According to Pinter et al. (2008),^[2] although climate change has made flooding more likely over that time, “the largest and most pervasive contributors to increased flooding on the Mississippi River system were wing dikes and related navigational structures, followed by progressive levee construction. In the area of the 2008 Upper Mississippi flood, for example, about 2 m of the flood crest is linked to navigational and flood‐control engineering. Systemwide, large increases in flood levels were documented at locations and at times of wing‐dike and levee construction.”

Figure \(\PageIndex{6}\): Wing Dykes on the Mississippi River

Wing dykes were constructed to enhance navigation on the Mississippi system (not to reduce flooding), but Pinter et al. found that they resulted in increases in water levels in upstream areas because the wing dykes acted like dams. Levees were constructed to control flooding, but Pinter et al. found that they led to increased flooding downstream because they resulted in a loss of water storage in flood plains.

Media Attributions

• Figure \(\PageIndex{1}\): Jet Stream, from the US National Weather Service, public domain, https://www.weather.gov/jetstream/refl
• Figure \(\PageIndex{2}\): Steven Earle, CC BY 4.0
• Figure \(\PageIndex{3}\): Support During Hurricane Harvey from South Carolina National Guard, 2017, public domain image via Wikimedia Commons, https://commons.wikimedia.org/wiki/F..._(TX)_(50).jpg
• Figure \(\PageIndex{4}\): (left) Riverfront Ave Calgary Flood by Ryan L. C. Quan, 2013, CC BY SA 3.0, via Wikipedia, https://en.Wikipedia.org/wiki/2013_A...Flood_2013.jpg; (right) Okotoks June 20 by “Stephanie N. Jones” (Jadelicia), 2013, CC BY SA 3.0, via Wikimedia Commons, https://en.Wikipedia.org/wiki/2013_A...yground-03.JPG
• Figure \(\PageIndex{5}\): Steven Earle, CC BY 4.0, based on Hydrology data from LMNO Engineering, Research and Software, http://www.lmnoeng.com/Hydrology/rational.htm
• Figure \(\PageIndex{6}\): River Training Structures from US Army Corps of Engineers, public domain, https://www.army.mil/article/70627/r...for_navigation

---

1. Gupta, V. & Sah, M. (2008). Impact of the Trans-Himalayan Landslide Lake Outburst Flood (LLOF) in the Satluj catchment, Himachal Pradesh, India. Natural Hazards 45, 379–390. https://doi.org/10.1007/s11069-007-9174-6 ↵
2. Pinter, N. et al. (2008). Flood trends and river engineering on the Mississippi River system. Geophysical Research Letters, 35(23). https://doi.org/10.1029/2008GL035987 ↵