Banqiao Dam
© Wikimedia Commons.An aerial view of the Banqiao Dam following its failure as a result of mainland China's greatest 24-hour rainfall, which took place in Henan Province during the passage of Typhoon Nina in early August 1975. As many as 230,000 perished as a result of the dam collapse.
In early August 1975, Typhoon Nina made landfall along the coast of China near Shanghai. The storm weakened as it moved inland and merged with a cold front that straddled Henan Province (in central China), resulting in the development of a series of near-stationary thunderstorms in and around the town of Linzhuang. The greatest short-term rainfall event in modern Chinese records ensued: some 1060 mm (41.73") of rain fell in just 24 hours. The Banqiao Dam on the Ru River near Linzhuang, failed as a result of the rains and collapsed on August 7-8, sending a wave some 6 miles wide and 10-23 feet deep across the plains below. As a result some 100,000-230,000 people perished. By some accounts, this makes Typhoon Nina the fourth-deadliest tropical storm in modern world history.

The wide range in fatality estimates is because many of the deaths occurred as a result of famine and disease following the flood itself. The exact figures for each source of fatalities remains obscure. If the high-end figure of 230,000 is correct, then the event would rank as not only the fourth-deadliest tropical storm on record but perhaps the sixth-deadliest natural disaster of any kind since 1900. The caveat to these kinds of listings, however, is that when tremendous numbers like this are involved there are huge discrepancies in the range of fatalities attributed to each event (as was the case with the Banqiao Dam disaster). These discrepancies can be seen in Wikipedia's list of the top 10 deadliest natural disasters since 1900. The list uses the highest estimate for the Banqiao Dam disaster.

The Banquao Dam was constructed with the help of Soviet consultants in 1951 as a project to control flooding and generate electricity. It was designed to survive a "once in 1000 years" return rainfall event (the area normally receives about 1000mm/40" of rainfall a year) which was calculated as being 300 mm (11.81") of rainfall over 24 hours. In reality, that much rain fell in just two hours during the 1975 storm, with following maximum precipitation point rainfalls measured:


Comment: We're seeing similar record rainfall patterns nowadays. From just this month:


189.5 mm (7.46") in 1 hour
494.6 mm (19.47") in 3 hours
830.0 mm (32.68") in 6 hours (world record)
954.4 mm (37.57") in 12 hours
1060 mm (41.73") in 24 hours
1629 mm (64.13") in 3 days

The construction and maintenance of the dam was clouded by controversy, with one of China's leading hydrologists, Chen Xing, who was involved in the dam's design and construction, concerned that not enough sluice gates had been installed. He was removed from the project. The website internationalrivers.org has a detailed history of the dam project and its failure. The disaster was a major embarrassment for the government, and some details were considered classified information until the year 2005. This is the reason that so few people understood the true nature of the calamity until just recently.

Maximum point rainfalls

One of the most critical issues facing hydrologists and engineers is calculating what the "worst case" scenario might be in terms of rainfall when they are tasked with the construction of dams and other infrastructure projects. In general they want to know what the heaviest rainfall for specific periods of time might be over the course of the projected life of the structure(s). In other words, if the dam is expected to stand without reconstruction for 100 or 200 years, then they must know what the maximum amount of rain might be in a "once in a 100- or 200-year" storm. These time spans are often called return periods or recurrence intervals; they are meant to convey probabilities and not to imply that such rains would be guaranteed to occur in those time frames.

In the U.S., the Hydrometeorological Design Studies Center at the NOAA/NWS Office of Water Prediction publishes tables of extreme maximum point precipitation totals for various time periods for both the U.S. and the world. Below is the latest incarnation of those tables (which are being updated at this time):
us rainfall
table maximum point precipitation
world precipitation map
world precipitation data
Figures 1 and 2. Tables of maximum point precipitation for the U.S. (top table) and the world (bottom table). Some of these reports are from unofficial sources and perhaps of dubious reliability, so the NOAA/NWS Hydrometeorological Design Studies Center (HDSC) is currently undergoing a review of them. See the HDSC website for larger versions of these maps and full data listings for the United States and the world.
How anthropogenic global warming may affect extreme precipitation events

As I mentioned above, these tables are currently under review (I am assisting). Pending verification, a new U.S. record for 24-hour rainfall may have just recently been set in Hawaii: 49.69" at Waipa, Kauai on April 14-15 last month. The HDSC uses data like this to help formulate precipitation frequency estimates and probable maximum precipitation amounts for each region and state in the U.S. See the HDSC website for this information. This has recently been a controversial subject, with many scientists attributing global warming with increased precipitation rates and consequent flooding events. As we have witnessed, several "once in 100 or 500 year" precipitation events have occurred in the same area (i.e. Houston, Texas and, just recently, Ellicott City, Maryland) in just the span of a decade or so.


Comment: AGW doesn't exist. Our planet is undergoing serious cooling with cold records being broken on every continent and, interestingly, this pattern of cooling and devastating floods was documented during the little ice age.


maryland flood
© Libby Solomon/The Baltimore Sun/AP.Figure 3. The historic town of Ellicott City, Maryland was deluged by 6" to 10" of rain in just a three- to six-hour period on the afternoon of May 27, 2018, resulting in a devastating flash flood. It was the second time in the just the past two years that the town endured such, the previous occasion being in July 2016. In both cases the rain event was characterized as a "once in a 1000-year" rainfall accumulation, although the extreme flooding was likely influenced by recent development upstream.
Herein lies the rub. How can useful predictions of the return event periodicity of extreme rainfalls be calculated accurately if the climate is changing so rapidly that the historical pattern no longer applies, or if development is changing how quickly storm runoff flows into watersheds? Jason Samenow analyzed the challenges in calculating return periods in a Capital Weather Gang article on May 29.

There are clear signs of a nationwide increase in extreme rainfall events. Chapter 7 of the 2017 USGCRP Climate Science Special Report, "Precipitation Change in the United States", posited the following, "Analyses of precipitation extreme changes over the United States by region (20-year return values of seasonal daily precipitation over 1948-2015, Figure 7.2) show statistically significant increases consistent with theoretical expectations and previous analyses. Further, a significant increase in the area affected by precipitation extremes over North America has also been detected. There is likely an anthropogenic influence on the upward trend in heavy precipitation, although models underestimate the magnitude of the trend."


Comment: It appears weather models are quite unprepared for the kind of weather we have in our future:


Another graphic in the report (Fig. 7.3) illustrates the incidence of 2-day precipitation events in the U.S. that exceed their 5-year recurrence intervals and how this has dramatically increased since the 1990s:
2 day precipitation events US heavy flood
© USGCRP Climate Science Special Report, Fig. 7.3.Figure 4. Index of the number of 2-day precipitation events exceeding the station-specific threshold for a 5-year recurrence interval in the contiguous United States, expressed as a percentage difference from the 1901-1960 mean. The annual values are averaged over 5-year periods, with the pentad label indicating the ending year of the period. Annual time series of the number of events are first calculated at individual stations. Next, the grid box time series are calculated as the average of all stations in the grid box. Finally, a national time series is calculated as the average of the grid box time series.
Should this trend continue, we may expect to see more extreme rainfall events occurring as time goes by. As the above report stated in its Key Findings section for U.S. precipitation change:

"Heavy precipitation events in most parts of the United States have increased in both intensity and frequency since 1901 (high confidence). There are important regional differences in trends, with the largest increases occurring in the northeastern United States (high confidence). In particular, mesoscale convective systems (organized clusters of thunderstorms)-the main mechanism for warm season precipitation in the central part of the United States-have increased in occurrence and precipitation amounts since 1979 (medium confidence)."
Christopher C. Burt
Weather Historian

The Weather Company's primary journalistic mission is to report on breaking weather news, the environment and the importance of science to our lives. This story does not necessarily represent the position of our parent company, IBM.

Christopher C. Burt Christopher C. Burt is the author of "Extreme Weather; A Guide and Record Book." He studied meteorology at the University of Wisconsin-Madison.