Analysis of flood conveyance capacity of small- and medium-sized river and flood managements

Recent years have witnessed great losses caused by floods due to the unknown of rivers’ flood conveyance capacities, which make it difficult for floods management. In particular, flood conveyance capacities of small- and medium-sized rivers are usually designed to meet the flood with a return period of 10 years but these rivers have not been paid enough attention. After years of scouring and siltation, the flood conveyance capacities of small- and medium-sized rivers are unknown, and flood disasters caused by reservoir discharge are unpredictable. The aim of this research is to set a standardized research process to find out the actual flood conveyance capacities of small- and medium-sized rivers. The hydrodynamic model is constructed and calibrated to evaluate the river’s flood conveyance capacity. The design flood data are used to make the simulation, and the flood conveyance capacity of the river channel is evaluated combined with the actual situation of the river. Then with water level observation points settled, the dynamic process of water level of the river channel can be represented and real-time assessment of flood risk of the river can be done. Taking the downstream of Yingna River, a typical small- and medium-sized river flowing into the ocean, as an example, the results showed that this standardized process can provide a reference for the analysis of flood conveyance capacity of small- and medium-sized rivers.


Introduction
Floods have been known as one of the most catastrophic phenomena (Alaghmand et al. 2012). About 1-2 percent land around the world suffers from floods regularly (Chen 1995), and according to the research on the economic loss caused by floods in China, the average loss per year during 1999-2019 is up to 15.5 billion dollars (Li 2019). Recently, the process of urbanization in the world changes the original geomorphology and the natural hydrological cycle, which has reduced the ground's ability of water storage capacity (Su et al. 2018). Meanwhile, the increasing of impermeable surfaces have decreased the drainage ability (Su et al. 2018). All these factors have aggravated the damages brought about by floods. So, there is more attention on the flood hazard mitigation. One of the most important factors to decrease the loss caused by flood disaster is the flood forecast and early warning, which is very important for flood control decision (Ni 2005). To make the flood control decision, the flood conveyance capacities of rivers must be estimated. The flood conveyance capacities of rivers can be an important guideline for the discharges of reservoirs, which lead it to be a key factor of a flood control decision. If the flood conveyance capacities of rivers are unknown, then once a large flood occurs, it is impossible to predict the impact of reservoir discharge on the downstream protection objects (Liu et al. 2008), and this may bring dangers to the cities downstream of the reservoir.
In recent years, many countries have strengthened the large rivers and their main tributaries. Many important rivers and large flood control junction projects have been reinforced and their flood control and security capacity have been effectively improved Ma et al. 2017). However, the management of small-and medium-sized rivers is generally lagging. Numerous small-and medium-sized rivers have not been treated uniformly and are still in the situation of suffering from floods if there are rainstorms . Even though many small-and medium-sized rivers' embankments have been repaired, only the flood discharge capacity of these rivers is known. And different parts of the same river have different designed values (Zhang and Bi 2014). After years, the conveyance capacities become unknown to people due to different kinds of changes along these rivers.
China has always been troubled by flood disasters. According to the statistics, in the 50 years during 1950 ~ 2000, the flood disasters in China have the trend of increasing both in spatial scale and magnitude (Li and Li 2004). And 62% cities experienced water-logging and flood catastrophes during 2008-2010, in which more than 100 cities experienced floods more than 3 times during this period in China (Su et al. 2018). Economic losses and non-economic losses have been caused by flood disasters (Braimah et al. 2014). For better flood managements, many researches about rivers' flood conveyance capacities have been done in China successfully using different kinds of hydrodynamic models, including HEC-HMS, HEC-RAS, MIKE11, SWMM and so on (Liu et al. 2007;Cai and Yin 2015;Feng 2016;Chai 2016;Tang 2018). However, insufficient attention has been given to small-and medium-sized rivers that lack actual measured data (Wang et al. 2021). At present, the frequent occurrence of safety problems in small-and medium-sized rivers has been seriously threatening the safety of people's lives and property (Cheng 2020). Some small-and medium-sized river basins have serious water and soil loss (Maltsev and Yermolaev 2020). In addition, unreasonable sand mining, blocking the river, dumping garbage into the river, illegal construction and other encroachments on the river are increasing. Desilting has not been implemented for many years, resulting in serious shrinkage of the river and gradual reduction of flood conveyance capacity . For the rivers flowing into oceans, tidal jacking at the estuary can be an important factor hindering river evolution. But in practice, the simulation research seldom considers the external influence conditions from the tide of estuaries (Miao 2018). All the above factors of small-and medium-sized rivers seriously threaten the safety of urban and rural flood control.
The purpose of this research is to settle a set of standardized process for studying the conveyance capacity of small-and medium-sized rivers lacking measured data. To begin with, the hydrodynamic model is constructed and calibrated based on the measured data and the historical data. Then, the design flood is calculated to be the boundary condition of the hydrodynamic model for simulation of the evolution of flood. And the flood conveyance capacity of the river can be evaluated by analyzing the simulation results combined with the actual situation of the river bank. Ultimately, several water level observation points are settled down to represent the dynamic water level process of the whole river, and the real-time assessment of flood risk can be conducted. In this research, a typical river was chosen to verify the effectiveness of this standardized process.

Study area
This research takes the downstream of Yingna River as a case study, which is a typical river among the small-and medium-sized rivers flowing into the ocean. Yingna River is located in the northeast of Dalian, Liaoning, China, with latitude of 39°43′~40°15′N and longitude of 122°51′~123°14′E (Fig. 1). The altitude of the river source is 609.6m. The total length of the river is 94.9km, and the total basin area is 884.1km 2 . Yingna River Reservoir, as an important water source for Dalian, is built on Yingna River, and below the reservoir is the downstream area where many people live. There are two tributaries in the downstream of Yingna River contributing a lot of discharge into the main channel: Xinfang and Gunzi River. The flood control capacity of Yingna River Reservoir is limited because its normal high water level (79.1m) is only 1 meter higher than its flood limit water level (78.1m), and the difference in storage capacity is only 23 million cubic meters (Dong 2016). This situation leads to great flood control pressure on the downstream river, which is caused by the discharge from the reservoir during flood season. Besides the limitation of flood control capacity of Yingna River Reservoir, the natural factors also contribute to the flood control problem. The temporal distribution of rainfall in Yingna River Basin is very uneven, which mainly concentrates in June to September (Wang 2016), accounting for 65%-75% of the whole year's rainfall (Gao and Liu 2011), and this leads to the frequently occurrence of floods in this basin. In general, floods with high peak height and large volume of 3 days (Gao and Wang 2012;Chen 2015) occur during rainstorms. The downstream main channel of Yingna River has not been leveled and repaired for many years, coupled with the years of scouring and silting, which have led to the substantial changes of the cross-sections' shapes, and this has an important impact on the flood carrying capacity of the river channel. In addition, the downstream river often dries up in the dry season, leading to the overgrowth of weeds in the channel. And the weeds contribute to the effect of hindering flowing of flood. In recent years, some new wading projects have been built along the river, which also has the effect of raising the water level of the river. Moreover, the tidal jacking effect at the estuary of Yingna River also affects the flood conveyance capacity, which will aggravate the flood disaster in extreme cases (Zhong et al. 2012;Lian et al. 2013Lian et al. , 2017Zheng et al. 2014).

Data
In order to construct the hydrodynamic model successfully, the data including crosssection data of downstream of Yingna River channel, underlying surface information of the downstream basin, history flood process and history tide level process as well as designed flood process and designed tide level process are indispensable. Cross-section data of Yingna River were collected for building the hydrodynamic model. The downstream cross-section data of Yingna River were collected from an engineering project, in which a professional survey team was employed in 2020 to gauge the river cross-sections, and the layout is shown in Fig. 2.
The underlying surface information of the basin was collected for analyzing the simulation results. The residential area around the river is provided by Google Earth software (Fig. 3).
Collected historical flood process data were used for parameter calibration of the hydrodynamic model. No.803 Flood in 2017 is the nearest flood hitherto, so its flood process whose quality is the best were collected and analyzed for calibration. With the help of Dalian Water Authority, four flood traces were collected, two (Point 2, Point 3) are the measured data in 2017 by water gauge, and the other two (Point 1, Point 4) were collected through site investigation, the locations of those 4 points are shown in Fig. 3. The discharge time series during No.803 Flood period of the reservoir was collected from Yingna River Reservoir Management Department. The tide level data during No.803 Flood of the tide station were collected for parameter calibration and the statistical data of multiple-year max tide level was collected for simulation calculation. All these tide data were collected from the Chinese Tide Network. Based on the situation that there is not sufficient measured rainfall data which should be collected for the determination of the magnitude of rainfall in the lower Yingna River Basin, the magnitude of the rainfall during No.803 Flood in the Fig. 2 The layout of gauging cross-sections of lower Yingna River 1 3 lower Yingna River Basin was identified based on the analyzing for No.803 Flood (Zhou 2018a). And because of the lack of measured multiple-year streamflow data, it's necessary to use the designed methods to calculate the designed flood of the river (Zhang and Tan 2016). This research uses the calculation method in Calculation Method of Designed Rainstorm Flood for Small-and Medium-sized Rivers (areas without data) in Liaoning Province (CDFSMR) to calculate the designed floods (Wang et al. 1998).

Methodology
The chosen model in this research, MIKE 11 hydraulic model, was developed by the DHI (Danish Hydraulic Institute) in 1987. And the hydrodynamic model (Mike 11HD), in which St. Venant equations are used to compute the space-time variation of flow and water level in the rivers (Tayfur et al. 1993;Wang and Ni 2000;Gottardi and Venutelli 2004), is the core of MIKE 11, which has been widely used for researches in water resources engineering (Yaroslav et al. 2000;Stephen Harding Blake 2001;Mishra et al. 2001;Hammersmark et al. 2005;Hashemi et al. 2008;Ranaee et al. 2009;Sun and Liu 2010;Kusre et al. 2010;Makungo et al. 2010;Luu et al. 2010). MIKE 11 river model adopts hydrodynamic model (HD model), which is the implicit finite difference solution of the unstable flow in the channel. The six-point central implicit difference scheme is used in the difference scheme, and the traditional "chasing method", namely "double sweep" algorithm, is used in the numerical calculation. The model can be utilized to simulate the flow conditions and subcritical flow in different regions, or vertical homogeneous flow conditions from steep mountainous rivers to tidal estuaries. In addition, various simplified flow simulations, such as diffusion wave, motion wave and quasi steady flow, can be carried out at the same time of complete hydrodynamic simulation. It's governing equations and discretization are shown as following. where Q is the discharge, A is the section area of the river channel, x and t are distance and time coordinates respectively, h is the river depth, C is Chezy coefficient, R is the hydraulic radius, g is acceleration of gravity, α is momentum correction factor.
The Abbott-Ionescu six-point finite difference scheme is used to discretize the Saint-Venant equations in MIKE 11. The discrete format is not used to calculate the water level and flow at each grid point at the same time but to calculate the water level or flow alternately in order, called h point and Q point respectively (Yang 1993;Zhou 1995;Thompson et al. 2004), as shown in Fig. 4.
The main steps in this research are shown in Fig. 4, as following: Firstly, the river's channel thalweg and cross-sections data were extracted from the gauged data. The Network File was established based on the river's thalweg which could define the geometric topological relationship of the river in the model. And the Cross-sections File was established based on the cross-sections data which would be used to define hydraulic conditions of the river's cross-sections. Then, HD (Hydrodynamic) Parameters File was established by giving the model initial water level conditions and roughness. Lastly, establishing blank Boundary data Files for subsequent input, this file is for defining the hydrology input process for the model, like discharge process, tide lever process and etc. So far, MIKE 11 hydrodynamic model was established.
Secondly, because of the lack of rainfall measured data, the inflow process from rainfall to Yingna River was got through calculating using the method described below for calculating designed flood process after determining the magnitude of No.803 Flood. Then the processes of inflow, water level and discharge of No.803 Flood were input into the Boundary data File to carry out the parameter calibration of MIKE 11 hydrodynamic model.
Thirdly, based on the condition that the designed value of the river channel is 10-year return period and considering the possible reduction of the flood conveyance capacity of the river channel, this study calculated the 10 and 20-year designed floods using the method descripted below. The tidal designed value was calculated using the method in Code for Design of Embankment Engineering (Mei et al. 2013). Then, considering the most likely combinations of floods and tides, the combinations of floods and designed tide levels were input into the Boundary data File for simulation.
Fourthly, the flood conveyance capacity of every cross-section of the river was determined after comparing the simulation results with the actual river's bank level. At specific cross-section, if the bank level is higher than the maximum simulation water level, then the cross-section can be considered as having a higher standard conveyance capacity, if not, then the cross-section is considered as having a lower standard conveyance capacity.
Finally, water level observation points were determined based on the principles that are presented in Fig. 5. And after figuring out the maximum conveyance capacity of the crosssection which is most likely to suffer from floods, the 10-year designed flood without the discharge of the reservoir was input into the Boundary data File for the simulation to get the flow process of the cross-section. Then by just adjusting the start time of the discharge process from the reservoir in the Boundary data File with the 10-year designed flood processes, their time relationship was fixed. The time relationship is that when the maximum discharge of the reservoir reaches the cross-section, the water surface profile caused by the flood is at the time of the fastest decline rate. This time relationship not only can make sure that the reservoir can discharge as planned, but also can achieve the best flow regulation effect.
The inflow of the downstream can be divided into five parts: discharge of Yingna River Reservoir, distributed inflow from Yingna River Reservoir to Xinfang River, Fig. 5 The six-point Abbott-Ionescu finite difference format alternate arrangement of water level and flow points confluence of Xinfang River, confluence of Gunzi River and distributed inflow from Gunzi River to the estuary.
Designed discharge process of Yingna River Reservoir was collected from Yingna River Reservoir office that the office is using. Because Yingna River belongs to the small-and medium-sized rivers without data in Liaoning Province, so the rest four designed flood processes are calculated using the method in CDFSMR (Wang et al. 1998), as shown below, and the characteristics of the four parts basins are figured out using ARCGIS 10.7.
Equation 3 is used to calculate the confluence time: where the values of x and y can be found in Table 3 − 1 of CDFSMR (Wang et al. 1998), L is the length of the river, J is the average slope of the river.
Precipitation of designed surface rainstorm is calculated using following formulas: where P T , P 24 , P 6 is the average value of annual maximum 3 days, 24 hours, 6 hours and 1 hour rainstorm respectively and C V is the corresponding skewness coefficient that can be found in Fig. 1-3 ~ 1-9, the point and surface reduction factor K F (A K F -t-F relationship line with 24h as a parameter can be used for checking the point and surface reduction factor introduced in CDFSMR) can be determined in Fig. 1-12 and the modulus coefficient K P can be determined in Table 1-1, P TPP 、P 24PP 、P 6PP 、P 1PP are the 3 days, 24 hours, 6 hours and 1 hour precipitation of designed surface rainstorm respectively with frequency P.
Attenuation index of rainstorm in 10min ~ 1h, 1h ~ 6h, 6h ~ 24h can be calculated using Eq. 8 to Eq. 10 Designed point rainfall of any duration P tP can be calculated using Eq. 11 to Eq. 13 When t ≤ 1h, (10) n 2P = 1 + 1.661Lg P 6P P 24P When 1 < t ≤ 6h, When 6 < t ≤ 24h, Rainfall intensity can be calculated using Eq. 14 The t in Eq. 11 to Eq. 14 is the confluence time, P Tpp represents precipitation of designed surface rainstorm mentioned in formula (4)  where W TP , W (T−24)P , W 24P are designed three-day flood volume, front peak flood volume, and main peak flood volume for different frequencies P respectively, φ P is the flood peak runoff coefficient for different frequencies P, α TP 、α (T−24)P are three-day flood volume runoff coefficient, front flood volume runoff coefficient for different frequencies P respectively, and F is the area of the basin. Then the flood hydrograph needs to be calculated. The shape factor can be calculated using following formulas: where γ denotes the shape factor of flood hydrograph. when γ > 0.05, a two peak generalized process line with γ as parameter and the total duration of 72h, whose front peak is triangular., is adopted.
when γ < 0.05, the designed peak flow Q P is adopted as the maximum flow, and the flood hydrograph is determined based on the flood volume (this situation is not involved in the calculation, so there is no introduction about it).
The main peak hydrograph can be determined by multiplying Q t /Q P which can be found in Table 3-3 (related to value of γ P ) in CDFSMR with Q P and adding base flow.
The front flood hydrograph of the basin in the east of Liaoning Province is a triangle of rising water for 9h and falling water for 12h (falling to the value of Q 21 ). Q 21 can be calculated using Eq. 20.

Model construction and calibration
The geomorphic change will affect conveyance capacity of downstream of channel (Slater et al. 2015;Guan et al. 2016). Improving flood resiliency in channel will require better understanding of rapid conveyance capacity changes (Li et al. 2020). Therefore, we use the latest data construction and calibration model to accurately reflect the conveyance capacity of downstream of Yingna River channel.
1D hydrodynamic was constructed through four steps:(1) Import gauged data of the river to form Network File; (2) Import cross-sections gauged data to form Cross-sections File; (3) Import the designed flow process to form Boundary data File; (4) Defining the Manning values of the river's channel to form HD Parameters File.
The article about analyzing No.803 Flood was used to judge the magnitude of No.803 Flood, which turned out to be a flood with a return period of 20 years (Zhou 2018b, a). Thus, the inflow process of the downstream of Yingna River during No.803 Flood is shown in Fig. 6.
Based on the underlying situation of the river channel, resistance value was set for each part of the river channel. The part near the dam was leveled several years ago, so its resistance value is lower than the ordinary river, which is between 0.015 ~ 0.029. The part near the estuary is not affected by weed due to the influence of tide, so its resistance value is on average, which is between 0.03 ~ 0.039. The rest part is affected by scouring, siltation and weed at the same time, so its resistance value is above average, which is between 0.04 ~ 0.055. All the above value ranges come from Hydraulic Calculation Manual (Li and Zhao 2006). Then after calibration, the resistance value of each part of the river channel is shown in Table 1.
The simulation results of the four points are shown in Table 2. It can be seen that the maximum difference between simulation results and the flood traces is less than 0.15m, which denotes the match effect is good. Due to the influence of tide level, point4 has a highest error. In addition, there are errors in the data of site investigation, it can also affect the error of simulation results. The hydrodynamic model can well reflect the real river flow evolution and can be used for simulation.

Conveyance capacity of downstream of Yingna River channel
According to the principle of the most likely occurrence (Zhou and Wang 2007), the combination of designed flood and designed tide level was selected to analyze the flood carrying capacity of the downstream of Yingna River. The combination of 10-year flood (Fig. 7) and 20-year high tide (2.99 m) as well as the combination of 20-year flood (Fig. 7) and multi-year average high tide (2.68 m) were chosen as the basis of analysis.  As for the parts with dike, the guaranteed water level is the designed flood level of the embankment or the highest water level beyond which flood disasters may occur, which is used to judge the flood conveyance capacity of the river.
As for the parts without dike, the flat bank water level is used to judge the flood conveyance capacity of the river.
If the guaranteed water level is higher than the simulated water surface elevation, then it is replaced by the simulated water level for the flood control safety because this would make people more sensitive to floods. The flood carrying capacity is shown in   Table 3, Table 4 and Table 5. Due to the different criteria of flood carrying capacity, the channel with dike is separated from the channel without dike. Table 3 and Table 4 show the simulation results of the left and right banks with embankments in the downstream channel of the Yingna River respectively. Table 5 shows the simulation results of downstream channel without embankment of Yingna River.The simulation results demonstrate that Sunjiabaozi can be the boundary of conveyance capacity of the downstream of Yingna River. According to the simulation results, the conveyance capacity of 76% cross-sections of the upper channel of Sunjiabaozi are greater than 20-year flood. The conveyance capacity of 16% cross-sections of the upper channel of Sunjiabaozi are smaller than 10-year flood, while 8% are great than 10-year flood but are smaller than 20-year flood. As the Google Earth shows, the upstream topography of Sunjiabaozi basin are mostly hilly and able to block floods, the downstream topography of Sunjiabaozi basin are mostly plains, which is conducive to flood spread. So the residents living in the upper Sunjiabaozi basin won't suffer from floods even when the 20-year flood comes ashore. As a result, the water conveyance capacity of the upper reaches of Sunjiabaozi can meet 20-year flood.
The conveyance capacity of 38% of the downstream channel of Sunjiabaozi meet 20-year flood. The conveyance capacity of 54% of the lower channel of Sunjiabaozi can't meet 10-year flood, while 8% meet 10-year flood but don't meet 20-year flood. The weak flood conveyance capacity reaches, around where people would suffer from flooding when 10-year flood comes ashore, are all located in the lower reaches of Sunjiabaozi. These reaches are all with low elevation on both banks, narrow width of the river channel, discontinuous embankment. Most of the places in the lower Sunjiabaozi basin where residents live on both banks are flat, and the residential houses are close to the river bank seen from Google Earth. All the above lead to the result that people who live in these places are vulnerable to floods. So this area is the key area for flood control in the downstream of Yingna River Basin.

Observation points
In practice, by selecting appropriate water level observation points, the water level dynamic process of the river in real time can be represented. According to standard for stage observation (GB T50138-2010), the selection principles of observation points are as follows: 1. Observation points should be selected in the places with easy to reach for transportation, easy to observe and close to town; 2. The observation points should be located in places where the river bank is straight, stable, not easy to be scoured or silted; 3. The distance between two observation points should not be too long so that the water level of the observation point can have a good effect of representing the general water level of the river.
Based on the principals, simulation results and the map, five water level observation points were settled for having a good representation of the dynamic process of the whole  Fig. 8. The characteristic water levels of the five observation points' sections are shown in Table 6. In order to ensure flood control safety, the lower guaranteed water level on both sides of the section is selected as the guaranteed water level of the observation point. And the dynamic process of water level of the parts of the river channel can be represented through the water level of the corresponding observation points.
When the water level of the observation point reaches the guaranteed level, nobody is allowed to close the shoreside, and the small bridges over the river are closed to traffic. Flood fighting materials should be used to strengthen and heighten the bank where danger may arise, and the residents should be evacuated to higher and safer places when the situation gets worse.

Suggestion for reservoir discharge
After analyzing the process of water level of the simulation results in Table 5, it turned out that the section of the overflow bridge near Caijia Village is the first place where the flood reaches the disaster line (24 hours after the beginning of floods). And this suggests that Caijia Village (close to the estuary of Yingna River) is most likely to suffer from flood disaster during flood season in the downstream of Yingna River. So it is necessary to find out the maximum discharge of Yingna River Reservoir during flood season under the condition that Caijia Village wouldn't be affected by floods.
The left bank elevation of overflow bridge section is 6.37m and the right bank elevation is 5.39m. According to the flood control demands of the bank (Zhou and Wang 2007), it may lead to a dangerous situation when the water level is 0.4m down from the bank elevation (which turns out to be 4.99m). Based on the results of hydrodynamic simulation, the water level of 5.04m of Caijia observation point matches with the water level of 4.99m  of the cross-section of the overflow bridge. When the flood peak discharge occurs at Caijia observation point (without the discharge of Yingna River Reservoir), it works well on adjusting peak flow in this point if the maximum discharge of the reservoir occurs 3 hours later according to the result of hydrodynamic simulation. Therefore, suggestion on reservoir discharge is shown in Table 7.
When the water level of Caijia Village observation point reaches 4.67m caused by the flood of 10-year return period, it is suggested that the maximum discharge of the reservoir is 979 m 3 /s, and then the highest water level of the observation point would be 5.04m.

Conclusions
The unknown of the downstream of Yingna River conveyance capacity has brought many troubles to its flood management. In this research, a set of standardized research process was used to find out the flood conveyance capacity of the downstream of Yingna River. Firstly, the 1D hydrodynamic model was constructed based on the downstream of Yingna River channel's gauged data in 2020. Secondly, the parameters have been calibrated with No.803 Flood related data (described above) which make this model reliable. Thirdly, the river flood and tide conditions were clarified through investigation. Then, after studying the actual most likely flood situation, the magnitude of floods and tides was combined to form calculation schemes and were input into the hydrodynamic model for simulation to obtain the river flood routing process. Fourthly, after comparing the guaranteed water level with the simulation result, the actual flood conveyance capacity of each section of the river was determined. Finally, based on the flood conveyance capacity of the river and the actual situation of the basin, several water level observation points were settled, then the dynamic water level of the whole river could be represented. And flood risk assessment could be implemented as well as flood control suggestions were provided.
The simulation results demonstrate that Sunjiabaozi can be the boundary of conveyance capacity of the downstream of Yingna River. The conveyance capacity of the upper channel of Sunjiabaozi can meet 20-year flood. The conveyance capacity of 54% of the lower channel of Sunjiabaozi can't meet 10-year flood, and people would suffer from flooding when 10-year flood comes ashore. In order to have a full representation of the dynamic process of water level of the whole downstream of Yiingna River, five observation points were settled. And corresponding flood control management can be taken according to the water level of these observation points. According to the simulation results, Caijia Village is the most prone place to flooding in the downstream of Yingna River. In order to ensure the residents not to be affected by floods, the discharge of the reservoir should not exceed 979 m 3 /s when the basin suffers from a 10-year flood based on the simulation results. These analyses about the flood conveyance capacity of the downstream of Yingna River have achieved good results, and the results will be helpful for the flood management. In conclusion, the standardized research process in this study can be well applied to other small-and medium-sized rivers lacking measured data.