Analysis on the ‘Jul.20’ extreme rainstorm and flood control countermeasures in Zhengzhou, China

In recent years, extreme weather and climate events caused by global climate change have become more frequent and intense, which results in a lot of serious social and economic loss, and the safety of people's lives and property are threatened. On July 20, 2021, Zhengzhou city, Henan Province, was hit by an extreme rainstorm, which caused an extremely serious urban waterlogging disaster. During the extreme rainstorm, 380 people died, and the direct economic losses reached CNY 45.9 billion (about USD 7 billion) (The State Council, Beijing, China. 7.20 Disaster Survey Team 2022; Xinhua News Agency, Beijing, China 2022). The rainstorm caused a lot of rivers to overflow (such as the Jinshui River and Jialu River) in Zhengzhou, and many places in the urban area were heavily flooded (Ran et al. 2021; Ren & Zhang 2022; Yin et al. 2022).

In recent years, studies on extreme rainstorms both at home and abroad have mainly focused on hydro-climatic trends and flood risk assessment. Darabi et al. (2023) examined the spatio-temporal climate variability over the Ceyhan River basin in southern Anatolia, Turkey by using historical rainfall and temperature observations recorded at 15 meteorological stations. Various statistical and geostatistical techniques were employed to determine the significance of trends for each climatic variable in the whole basin and its three sub-regions. Adu-Prah et al. (2019) studied the spatio-temporal evidence of recent climate variability in Ghana. Several extreme rainstorms were analyzed. Danandeh Mehr et al. (2021) carried out innovative and successive average trend analyses of temperature and precipitation in Osijek, Croatia. A new trend of heavy rainstorms was put forward. Ghaderpour et al. (2023a, 2023b) studied climate and river flow in the Athabasca River Basin (ARB) with the method of least-squares triple cross-wavelet and multivariate regression to investigate possible relationships between river flow, temperature, and precipitation in the time–frequency domain. The ARB in Canada is selected as a case study to investigate such relationships. Ghaderpour et al. (2023a, 2023b) analyzed precipitation time series for Italian Regions. In Italy, most of the destructive landslides are triggered by rainfall, particularly in central Italy. Effective monitoring of rainfall is crucial in hazard management and ecosystem assessment. In the research, the available monthly GPM data were employed to estimate the monthly precipitation for the 20 administrative regions of Italy from June 2000 to June 2021. Farooq et al. (2023) studied the flood forecasting model using federated learning. In this article, a flood forecasting model using a federated learning technique has been proposed. The proposed flood forecasting model has successfully predicted previous floods that happened in the selected zone from 2010 to 2015 with 84% accuracy. Todaro et al. (2022) analyzed climate change over the Mediterranean Region. The climate is changing and is resulting in heavy rainstorms and uneven distributions. Mersin et al. (2022) analyzed historical trends associated with annual temperature and precipitation in the Aegean Region, Turkey to find correlations between annual temperatures and rainstorms. Rosmanna et al. (2015) compared trends in hydrometeorological averages and extreme data sets around the world at different time scales. More than 50 annual data sets were analyzed to find extreme rainstorm trends. Hefzul Bari et al. (2016) analyzed seasonal and annual rainfall trends in the northern region of Bangladesh. The trends reflected fluctuation during recent years.

Although scholars all over the world have carried out a large number of studies on extreme rainstorms (Dunn et al. 2012; Li et al. 2014; Dams et al. 2015; Liu et al. 2023), there are still some problems (Lu et al. 2020; Abu Hammad et al. 2022; Jiang et al. 2022; Wijeratne et al. 2023). The first is that the existing research results are often based on normal rainstorm data series, which lacks treatment methods for the extreme values, and the conclusions deviate greatly from the actual values. The second is that the existing studies lack a multiple flood disaster mechanism analysis of natural and human factors. The third is the uncertain hydrology problems while modeling flood events in urban areas, which is often the result of errors in flood simulation (Cappato et al. 2022). The fourth is that few studies deal with extreme urban flood frequency analysis methods and flood disaster management countermeasures.

Therefore, the main objectives of this paper are (1) to propose a rainstorm long-series analysis method, including rainstorm anomaly analysis method, and construct flood frequency calculation models; (2) to calculate the rainstorm and climate factor anomalies by using the studied methods based on 72 years (1951–2021) of rainstorm records and to reveal the extreme rainstorm deviations compared with the mean values in different periods in Zhengzhou; (3) to select the short-duration rainstorm data from Jul.20 and calculate the rainstorm frequency in different periods; (4) to analyze the Jul.20 extreme rainstorm characteristics and formation mechanisms; (5) to calculate the urban flood depth and draw the flood inundation distribution map; and (6) to put forward urban flood control countermeasures in view of extreme rainstorms and flooding.

The studied results have an important reference value both for extreme rainstorm prevention and urban flood disaster alleviation.

The terrain of Henan Province is high in the west and low in the east, and the Taihang, Funiu, Tongbai and Dabie Mountains are distributed in the north, west and south of the province, with a semi-circular shape along the provincial boundary. The central and eastern parts of the province are the Huang-Huai-Hai alluvial plain, and the southwest part is the Nanyang Basin.

Zhengzhou is the capital city of Henan Province, located between 112°42′E–114°14′E, and 36°16′ N–34°58′N. Zhengzhou is located in the transition zone from the Huang-huai-hai plain in the eastern part to the mountainous area in the western part of Henan Province, with Songshan Mountain in the west and the Yellow River in the north. The city area is mainly plain. The total area accounts for 7,446.2 km². Zhengzhou is one of the 15 national central cities in China.

The Yellow River flows through Zhengzhou. The main tributaries in Zhengzhou are the Jialu River, Shuangji River, Yinghe River and Qili River. The drainage channel and natural water storage area are relatively small.

In recent years, the extreme rainstorms in Zhengzhou has shown an increasing trend, the rainstorm frequency and intensity also show an increasing trend, and the rainstorm duration shows a downward trend, which shows that the extreme nature of the rainstorm in the Zhengzhou area is gradually increasing.

The annual rainstorm and meteorological data come from 247 national meteorological stations from January 1951 to July 2021 (Zheng et al. 2023). The short-duration data for the Jul.20 rainstorm (5 min, 10 min, 15 min, 20 min, 30 min, 45 min, 60 min, 90 min, 120 min, 150 min, 180 min, 240 min, 360 min, 540 min, 720 min, 1,440 min, and 3 days) come from 18 national meteorological stations in Zhengzhou.

The above data series were verified and tested by the National Meteorological Information Center.

The radar map data comes from the PUP product data uploaded by the National Weather Radar Stations. The horizontal radar map resolution is 0.01° × 0.01°, and the temporal resolution is 6 min.

The fifth global climate reanalysis data set (the fifth global reanalysis, ERA-5) issued by the European Medium-term Weather Forecast Center was used to analyze the atmosphere circulation during the Jul.20 extreme rainstorm. The data can be updated in real time, and high-resolution global atmospheric reanalysis data from 1951 to 2021 were provided and released in a global grid format. The horizontal resolution is 0.25° × 0.25°, and the temporal resolution is 1 h. The ERA-5 reanalysis data set is oriented and divided vertically as 37 layers from 1,000 to 1 hPa.

The flood disaster and urban inundated data as well as flood loss statistic data come from Xinhua News Agency (2022) and the State Council Jul.20 Disaster investigation team (2022).

Here, is a rainstorm variable (mm); is the mean value in the calculation period (mm).

Due to that, the water vapor in the atmosphere is mainly concentrated in the lower layer of the troposphere, the whole layer integration of the water vapor flux in this paper refers to the vertical integration from the Earth's surface to the height of 300 hPa.

Here, x is the rainfall variable; is the location parameter; is scale parameter is the shape parameter. When ξ < 0, it belongs to the Weibull distribution, when ξ = 0, it belongs to the Gumbel distribution, when ξ > 0, it belongs to the Frechet distribution.

When the three parameters of the GEV distribution are obtained, we can carry out the goodness-of-fit test for GEV by using the K–S test. The parameters meaning in the above formula is the same as the formula (5).

Substituting the maximum likelihood estimation parameters into the formula (7), the flood extreme value X_p corresponding to the theoretical recurrence period T can be obtained.

According to the characteristics of the Jul.20 extreme rainstorm, the maximum rainstorm volumes for 17 series (5,10, 15, 20, 30, 45, 60, 90, 120, 150, 180, 240, 360, 540, 720, 1,440 min, and 3 days) were retrieved from the Zhengzhou Meteorological Station, as shown in Table 1.

The historical rainstorm data series (1951–2021) in this paper were collected from the Zhengzhou Meteorological Stations. The comparison of the calculated results between the Jul.20 and historical rainstorm is shown in Table 2.

The study results show that, during the Jul.20 rainstorm process, the frequency of maximum rainstorms in all of the different short durations is greatly over 1/1,000.

From 17 to 18, July, the rainstorm first started from the north part (Jiaozuo, Xinxiang, Hebi and Anyang region) of Henan Province. From 19 to 20 July, the rainstorm strengthened, and the rainstorm center moved toward the south direction. Then, an extreme and long-duration rainstorm appeared in Zhengzhou. The rainstorm process showed a significant multi-peak structure (Wang et al. 2022).

A weak pressure ridge appeared in the northern part of Henan Province. In the eastern Yellow Sea region, there has been a high-altitude cold vortex since July 19. The Japan Sea has an abnormally strong and obstructed high-pressure ridge. Usually, the airflow at 200 hPa potential height field is relatively flat and straight. However, the Jul.20 rainstorm process presented a very high-altitude wave column, and the phase was stable and less active during the occurrence and development of the Jul.20 extreme rainstorm. Henan Province is located in front of the abnormally strong high-altitude trough, and the high-altitude air flow is strongly scattered, the divergence degree at 200 hPa exceeds more than 4 times the climatic mean standard deviation, which provides favorable conditions for the development of the underlying convection system. As for 500 hPa, the subtropical high pressure was abnormally strong and located northward in the early stage of the Jul.20 extreme rainstorm process. The ridge suddenly jumped northward on July 15 and remained steadily at the north region of 37° N, which has 13 latitudes deviation compared with the normal annual mean. It is shown from the high-altitude map at 500 hPa on July 20 (Figure 4(b)) that the western Pacific subtropical high pressure is significantly strong in the northeast region of Henan Province. An eastward wind water vapor transport channel was established on the south side. It is worth noticing that the abnormally strong and northward deviation degree of the subtropical high pressure was more than twice that of the climatic average standard deviation.

The northwest part of Henan Province was controlled by an unusually strong continental high pressure, and there was a small-scale vortex in the western region of Henan Province. At the same time, Typhoon No.6 (Fireworks), located in the northwest Pacific Ocean, was constantly moving northward and strengthening, The eastward air flow in the north region was also strengthened. Henan Province was located between the abnormally northward strong subtropical high pressure and the typhoon ‘Fireworks’ from the Northwest Pacific Ocean. The combined strong atmospheric pressure gradient was generated and the eastward airflow was also strengthened, which provided an abundant source of water vapor for the Jul.20 extreme rainstorm process. In addition, typhoon No. 7 (Chapaka), located in the South China Sea, was also strengthened at the night of July 19 and continued to move northward. On 850 hPa in the low level (Figure 4(c)), significant cyclonic shear began to appear in western Henan province. It is worth noting that on the circulation characteristic map on 925 hPa (Figure 4(d)), there were three air streams to converge into the central Henan Province. The first was the southeast airflow in the south, the second was the eastward airflow in the east, and the third was the northeast airflow that was blocked by the Taihang Mountain and turned to the south direction. Three air streams carrying a large amount of water vapor converged in the Zhengzhou region, which is an important reason for the occurrence of the Jul.20 extreme rainstorm.

Under the favorable circulation configuration, the abundant water vapor was continuously transported to the rainstorm area along the channel and converged in the Zhengzhou region, which resulted in strong rainstorm intensity and long-duration rainstorm process.

Due to strong northward subtropical high pressure, and the eastward airflow on the southern edge being superimposed by the eastward airflow on the north side of typhoon ‘Fireworks’, a very strong water vapor transport channel from east to west was established, which continuously transported the water vapor from the northwest Pacific Ocean to Henan region. Besides, Typhoon ‘Chapaka’ in the South China Sea was strengthened, continuously moved to the north direction, and approached Guangdong Province. Under the combined action of ‘Chapaka’ and the strong southwest monsoon, a water vapor channel from south to north was also established, which continuously transported the water vapor from the south to the Henan region. Two water vapor channels converged in Henan Province, making the average water vapor flux in the central and eastern parts of Henan Province reach more than 600 kg/(m.s), which exceeds three standard deviations from the normal climatic state. It shows that under the co-influence of multiple large-scale systems, the water vapor transport in Zhengzhou was extremely strong and reached extremely favorable conditions for Jul.20 extreme rainstorm occurrence.

The short-duration rainstorm in Zhengzhou shows the obvious feature of localities. The hourly rainstorms were significantly higher in Zhengzhou than that in the surrounding areas, which is closely related to the topography. The landform and topography in the Zhengzhou area are relatively complex, spanning the second and third landform steps of China. The topography goes from low mountain, and middle high mountain, hills to plain, mountain hills. The plain boundary is very obvious.

Zhengzhou region is located in the northeast of Funiu and Songshan Mountain, which are situated in the south of the Taihang Mountains. There is a strong eastward torrent in eastern Henan, which caused water vapor to converge to Zhengzhou, and stretch to the Songshan Mountains in the west of Zhengzhou. When this strong eastward torrent encounters the Taihang Mountain, due to the topography block, it will be diverted to the Zhengzhou area. At the same time, in the south region of Zhengzhou, there is another southward airflow moving to the north along the Funiu Mountain. There is a very low altitude air flow at the same time in the south, east and north parts around Zhengzhou, The west was blocked by the mountains, which provides very favorable conditions for airflow rising movement and water vapor convergence in the Zhengzhou area.

Therefore, under the stable circulation situation, the eastward and southward airflow carrying a large amount of water vapor converged in Zhengzhou to cause the Jul.20 extreme rainstorm. Generally, the upward movement of the rainstorm is mesoscale, but during this special rainstorm process, a strong upward movement occurs in a large area near the rainstorm area, an important reason is topography block and uplifting action around (the northern, western and southern) Zhengzhou, which puts Zhengzhou region in the strongest area of upward movement and low-layer convergence. The airflow vertical velocity at low altitude 850–700 hPa was very obvious, and the maximum standard deviation was more than 8 compared with the normal climate state, which indicates that the dynamic uplift role is more significant than the water vapor role. Under the cooperation of abnormally strong convergence action at the high layer, the Jul.20 extreme rainstorm appeared in this area, which indicates that the topography in the northern, western and southern parts around Zhengzhou played an important role in the Jul.20 extreme rainstorm process.

The circulation of urban heat islands causes the rise of air convergence and uplifting, coupled with more condensation nuclei over the city than over the suburb, which eventually leads to more rainfall over the urban and downwind area than over the suburb. This effect is called the rain island effect.

Hu et al. (2023) carried out the integrated evaluation and quantitative research on the urban rain island effect of Zhengzhou with multi-methods. The social and economic data of Zhengzhou from 1978 to 2018 and the daily rainfall data from 1960 to 2020 were selected, and the evaluation method system of the rain island effect was constructed by multiple methods. The results show that the rain island effect in the central area of Zhengzhou has an obvious rising trend from 1997 to 2025. The increasing trend of the annual and seasonal rainfall island effect will be maintained for 8 and 20 years, respectively. The first main cycle of the rainfall island effect is 8–10 years and the second main cycle of the rainfall island effect is 17–20 years. The annual and flood season rainfall anomalies calculated by using the standardized anomaly method were 1.261 and 1.643, respectively.

For comparison with the Jul.20 extreme rainstorm, 24-h rainfall of different frequencies in the districts was calculated based on the data series from 1951 to 2021, as shown in Table 3.

Most water logging points in the Zhengzhou urban area are generally situated in the Jinshui district. There are several rainfall observation stations distributed in the Jinshui district. Therefore, we selected a typical area in the Jinshui District in this study to carry out a detailed study. The selected study area is 64.42 km².

The Jul.20 extreme rainstorm weather process was characterized by long duration, large cumulative rainfall and extreme rainfall intensities, which led to extremely serious flood disasters.

The circulation of urban heat islands causes the uplift of air convergence, and the rainfall effect of urban heat islands is significant.

• The construction and development of drainage pipe networks are unbalanced. By the end of 2020, the total length of the main drainage pipe in Zhengzhou was 2,456.8 km, which was mainly concentrated in the new urban areas. The upgrading and renovation of the pipe network in the old urban areas is seriously lagging behind.

• Rainwater regulation and storage planning were disconnected from the construction and operation. By the end of 2020, 31 rainwater regulation and storage facilities have been built in Zhengzhou, with a total regulation and storage capacity of 3.079 million m³. The regulation and storage facilities are not consistent with the planning and have no effect on the goal of storage and discharge balance within the planning scope.

• Flood discharge channel construction and pumping station capacity did not meet the relevant standards (Ministry of Housing and Urban-Rural Development, Beijing, China 2017). At present, 40% of the urban drainage pumping stations cannot meet the standard. By the end of 2020, the number, length and designed overflow capacity of the roads in the main urban area cannot meet the planned target.

• The management system of the drainage project is complex and there is insufficient overall coordination. The 10 rivers in Zhengzhou urban area are managed by different departments and units. The upper and lower reaches of a river are separately managed by different departments, which causes multi-levels and difficult coordination. The departments and administrative units are independent of each other, and there is no unified management department for river courses in the whole city. For example, the management of municipal drainage and waterlogging prevention facilities (urban pipe networks, open ditch, etc.) are separately under the leadership of 18 different units and 22 management responsibility sections.

• Urban flood control emergency plans need to be improved. The existing plans (Zhengzhou Municipal Flood Control Office, Zhengzhou, China. 2018) are obviously insufficient in the aspects of early risk perception and prevention measures.

• It is very often that the relevant responsible departments would begin to take measures when the flood disaster is very serious. It is short of beforehand measures for flood disaster prevention. In addition, the flood control plans in many district levels are not closely combined with the actual situation, Some water management units have problems such as unspecific flood control plans and inoperable flood countermeasures.

It can be seen from Table 2 that according to the calculation results of the historical extreme rainstorm frequency, the maximum rainstorm volume in 10 min relative to 1/1,000 frequency is 31.6 mm. For the Jul.20 extreme rainstorm, the maximum rainstorm volume in 10 min appeared from 16: 40–16: 45, July 20, during which the rainstorm volume reached 33.7 mm and the relative frequency was over 1/1,000. The maximum historical rainstorm volume in 60 min (1 h) for 1/1,000 frequency is 110.3 mm, and for the Jul.20 rainstorm, the maximum rainstorm in 60 min (1 h) appeared from 16: 00–17: 00, July 20, during which the maximum rainstorm volume reached 201.9 mm and the frequency is greatly over 1/1,000. The maximum historical rainstorm volume in 1,440 min (24 h) for 1/1,000 frequency is 216.8 mm, and for the Jul.20 rainstorm, the maximum rainstorm in 1,440 min (24 h) appeared from 20: 00 July 19 to 20: 00, July 20, during which the maximum rainstorm volume reached 552.5 mm and the relative frequency is greatly over 1/1,000. For historical rainstorms, the maximum rainstorm in 3 days relative to the 1/1,000 frequency is 455.3 mm. In the Jul.20 rainstorm, the maximum rainstorm in 3 days is reached to 624.1 mm, which is greatly over 1/1,000 frequency.

Considering climate change during recent years, some countermeasures should be improved to face the future climate change situation.

• Improvement of water logging prevention and control standards.

In the process of extreme rainstorms, it is necessary not only to raise the threshold of ‘no water accumulation’ and ‘no water logging’, but also to ensure the flood control safety of the river embankment in the upper reaches of the city. The design water level at the points between the urban pipe networks and the drainage open channels should be the river channel flood level under the corresponding extreme rainstorm situation, and should not simply assume the normal operation water level in the river.

• Improvement of river courses and riparian management.

In the past 30 years, the construction land in Zhengzhou has increased from 13.43% to 58.50%, which greatly changed the water system pattern. In urban spatial planning, a landscape pattern connecting rivers and green space should be formed. The river network stability should be maintained while realizing river ecological protection by using the existing woodland and water surfaces.

• The urban flood intelligence regulation and emergency command decision-making should be strengthened.

Intelligent tools are a development trend for urban flood regulation at present and in the future, and it is urgently necessary to improve flood disaster prevention. Although the Jul.20 flood disaster in Zhengzhou was caused by the extreme weather, there are also some problems and deficiencies in the emergency command operation, which could be called both a ‘natural disaster’ and a ‘human disaster’. The response procedures of the emergency departments should be further integrated, and the responsibility of the emergency departments for flood control should be clarified, so as to achieve joint prevention and control.

• The water vapor condition was an important factor in causing Jul.20 extreme rainstorm. Under the combined action of subtropical high pressure, typhoons ‘fireworks’ and ‘chapaka’, the water vapor flux through the whole vertical air layer in the rainstorm area greatly exceeded the historical climate average values. The anomaly degree reached more than three standard deviations. The water vapor flux divergence anomalies in the rainstorm area are very obvious. The water vapor convergence anomalies reached more than five standard deviations, which resulted in abundant water vapor flux conditions in the rainstorm area.

• The topography played an important role in the ‘Jul.20’ rainstorm in Zhengzhou. Under the obstruction and uplift action of Funiu Mountain, Songshan Mountain and Taihang Mountain, the rainstorm area has strong convergence and upward movement conditions in the lower-level horizontal wind field. The surrounding topography around Zhengzhou plays an important role in the stable maintenance of the rainstorm position and the increase of rainstorms.

• Rainstorm anomaly analysis methods were put forward in this paper to calculate different climatic factor anomalies. Flood frequency calculation models were constructed including the GEV distribution model and rainstorm frequency calculation models. Based on the above methods, the weather system anomalies and rainstorm frequencies have been calculated and analyzed.

• Taking the 72-year rainstorm statistic data as samples (January 1951 to July 2021), the rainstorm recurrence period of the Jul.20 was calculated and the flooded map was drawn. The calculation results show that the recurrence period of the Jul.20 extreme rainstorm in Zhengzhou is more than 1/1,000 years.

• The influence factors on the Zhengzhou Jul.20 flood disaster were analyzed, including natural (extreme rainstorm, urban rainfall island, etc.), topography, infrastructure as well as human factors.

• Based on the features of the Jul.20 extreme rainstorm, the countermeasures to raise the response-ability for extreme rainstorms are put forward, which include improving waterlogging prevention and control standards, strengthening river channel and riparian planning and management, realizing intelligent urban flood regulation and emergency command and decision-making.

This article does not contain any studies performed by other authors and meets the ethical standards.

The author consents to participate in the works under the Ethical Approval and Compliance with Ethical Standards.

All the data in the paper can be published without any competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

The study conception and design, material preparation, data collection, and analysis were performed by K.Z.

All relevant data are included in the paper or its Supplementary Information.

The authors declare there is no conflict.