KMS Institute of soil and water conservation Chinese Academy of Sciences
|Place of Conferral||北京|
|Keyword||海绵城市 洪涝灾害 水文水动力模型 效应分析 归因分析 萍乡市 江西省|
（2）系统分析了江西省萍乡城市建设驱动下的土地利用/覆被变化规律，构建了模型运行的气象、水文、土地利用/覆被等数据库，验证了海绵城市建设背景下分布式城市水文水动力模型的模拟效果。选择五丰河流域为研究区，基于2016-2018年高分影像和海绵设施现场调查数据，利用ENVI5.3平台对多源数据进行解译和像元叠加，构建流域土地利用/覆被栅格数据。分析了土地利用变化转移面积、概率矩阵和动态度，明确了在所有土地利用类型/覆被变化中海绵体历年变化最剧烈。其中，2017年增加了19 km2，动态度为25441%，2018年增加了77 km2，动态度为401%，增加面积均主要由广场和公路转化而来。根据土地利用变化规律，建立历年场次雨洪、土地利用/覆被、数字高程模型（DEM）、流向和累积汇流量等数据库。选择1小时和30米作为模型的时空尺度以满足模拟的准确性和时空异质性要求。采用反距离权重方法内插出网格逐小时雨洪过程、CN值、稳渗率值、地表糙率值等产汇流参数，确定坡面累积汇流量等级阈值为800，逐级进行坡面水动力过程演算。基于2016-2018年的模型数据库，通过14场洪水过程验证了分布式城市水文水动力模型模拟值与实测值间纳什系数的均值为0.68，表明模型模拟效果良好。
（3）评价了海绵城市建设措施对五丰河流域海绵技术改造小区-区域尺度产汇流的影响，揭示了影响的成因。根据五丰河流域土地利用/覆被变化规律，分别将2016年和2017、2018年分别作为效应评价基准期和分析期，利用分布式城市水文水动力模型，分别模拟了基准期和分析期产汇流过程。模拟分析表明：经过海绵城市建设，五丰河流域CN值减小了31%，各不透水铺装地面CN值平均降幅达到61%。CN值减小导致产流能力的降低，总量分别为20-30 mm、30-40 mm、40-90 mm和90 mm以上场次暴雨相应径流系数平均减少了44%。其次，五丰河流域稳渗率增加了53%，道路和公共管理用地稳渗率增幅超过20倍。稳渗率变大导致地表净雨量占净雨总量比例降低，地下净雨量占比提高。地表净雨量占比减少，表明地表产流机制相对于地下被削弱，与场次径流系数减少的影响叠加，造成地表径流系数减少，地表产流能力降低。地下净雨量占比增加，表明地下产流机制相对于地表加强，但仍不足以抵消径流系数减少的影响，导致地下径流系数减少，地下产流能力降低。此外，五丰河流域地表糙率增加了27%，各不透水铺装地面地表糙率增加了3倍，而马斯京根蓄量常数增加了91%。诸项汇流参数的变化导致地表洪水单宽流量减少了37%，表明流域汇流过程被阻滞。海绵设施在改造小区尺度发挥源头削减、过程阻滞和末端调控功能。其在流域上犹如分布式“水库”，增强流域径流控制能力，抑制地表超渗产流机制，加强地下产流机制，阻滞汇流过程。
（4）评价了五丰河流域海绵城市建设措施对洪水的影响，分析了洪水要素对产汇流参数的敏感性，揭示了影响的成因。经过海绵城市建设，五丰河流域不同量级暴雨（总量分别为20-30 mm、30-40 mm、40-90 mm和90 mm以上）所形成洪水的洪峰流量平均减少了38%，洪水总量平均减少了14%，洪水过程坦化。设置CN值、稳渗率值和地表糙率值3个因素和5个水平，采用正交试验制定敏感性试验方案。通过模型模拟各因素不同水平下五丰河流域出口断面处形成的洪水过程表明，海绵设施源头削减和过程阻滞功能导致CN值降低、地表稳渗率和糙率增加，是洪峰流量、洪水总量减少的成因。其中，CN值降低呈现出更为明显的敏感性和边际效应，是洪峰流量、洪水总量减少的主因。
（5）定量评价了海绵城市建设前后城市内涝灾害的变化，揭示了城市内涝灾害对海绵设施的响应及其成因。经过海绵城市建设，五丰河流域不同量级暴雨（总量分别为20-30 mm、30-40 mm、40-90 mm和90 mm以上）产生的内涝积水深度时程均值减少了38%，内涝积水深度时程最大值减少了47%，内涝灾害面积缩小了51%，内涝灾害累积时长缩短了62%。究其原因，海绵设施源头削减功能导致地表净雨量减少，内涝积水深度、灾害面积、累积时长随之减少；过程阻滞功能导致地表糙率增加，内涝积水深度、灾害面积、累积时长随之增加。海绵设施源头削减功能导致的内涝积水深度、灾害面积、累积时长减少量大于过程阻滞功能导致的增加量。海绵设施的源头削减功能是内涝积水深度、灾害面积、累积时长减少的主因。
The urbanization of population is close to 60%, which will sustainably increase in the future. And the forest, grass, and wetland are replaced by the impervious pavement, leading the interruption of the surface and underground runoff. Accordingly, the capacity of the urban underlying surface declines to regulate and store the flood and waterlogging, which makes it happen constantly. The sponge city construction aims to restore the nature hydrological cycle and the capacity of regulation and storage. But the hydrological and hydrodynamic processes are complicated and characterized by the spatio-temporal variability between the sponge city construction measures and the flood and waterlogging. The runoff generation and concentration law need to be excavated using suitable model in the processes. Based on this law, the scientific planning and construction has been the heated and important issue to avoid the deterioration of the flood and waterlogging lead by the urbanization. This paper built the distributed urban hydrological and hydrodynamic model, which was applied to analyze the effect and reason of the sponge city construction on the flood and waterlogging regulation and storage in Pingxiang city, Jiangxi Province. The main results of this study were shown as follows:
(1) The functional requirement and principle of the model was analyzed to develop the distributed urban hydrological and hydrodynamic model. It was programmed to form the independent intellectual technology. Specially, the functional requirement of the model was set to simulate the runoff generation and concentration of the sponge facilities and the flood and waterlogging coupling processes in the sponge facilities-plot-region scales. According to the requirement, the model distributed urban hydrological and hydrodynamic model was built. The runoff generation module adopted the soil conservation service curve number (SCS-CN) model and excess infiltration model to characterize the relationship between the storm and flood. Moreover, the runoff concentration module adopted the hydrodynamic models to characterize spatio-temporal distribution of the flood and waterlogging. The overland, river, and underground flow concentration module adopted the kinematic wave equation of Saint-Venant equation, Muskingum method and recession curves, respectively. Then the overland flow concentration was implemented according to the orders of the accumulated flow calculated by D8 algorithm. Meanwhile, the overland and underground net precipitation were the water balance elements for the flow concentration in the 2-D computational grids. The sponge facilities influenced the runoff generation and concentration by altering the parameter CN, constant filtration rate, and surface roughness. Lastly, the model was interactively programmed by the WebGIS and IntelliJ IDEA to form the independent intellectual technology.
(2) The sponge city construction derived land use and cover law was systematically analyzed to build the hydro-meteorological and land use and cover database for the model. It was evaluated in the Pingxiang city, Jiangxi Province. Specially, the Wufeng River basin was selected as the study area and its high-resolution image and field investigation data were collected between 2016 to 2018. Those two-type data were overlaid to build land use and cover database of the basin by ENVI5.3 software. The dynamic degree, transition matric and probability matrix were calculated for the area of the land use and cover to prove the sponge facilities as the main land use and cover change. Its area increased by 19 km2 in 2017 and 77 km2 in 2018 with the dynamic degrees of 25441% and 401%, respectively. The sponge facilities were mainly converted from the square and road. Then the storm, flood, land use and cover, digital elevation model (DEM), flow velocity, flow direction, and accumulated flow databases were built. The temporal and spatial scales were confirmed to be 1 hour and 30 m to meet the accuracy and spatio-temporal variability, respectively. The storm, flood, CN, constant filtration rate, and surface roughness were interpolated to the 1-hour scale in the computational grids. The overland flow concentration was implemented according to the orders until 800 of the accumulated flow calculated by D8 algorithm. Lastly, the model was validated by the Nash-Sutcliffe efficiency coefficients based on 14 storms and floods. The mean coefficients were 0.68, which proved the practicability of the model.
(3) The response and reason of the sponge city construction on the runoff generation and concentration was evaluated in the plot-region scale. In special, the 2016 and 2017, 2018 were selected as the base and benefit analysis periods according to the land use and cover change. The runoff generation and concentration were simulated in those periods by the model. The CN decreased by 31% after the sponge city construction in the Wufeng River basin and by 61% for the permeable pavements. It indicated the decrease in the runoff generation capacity. Actually, the runoff coefficient decreased by 44% for the storms of 20-30 mm, 30-40 mm, 40-90 mm, and above 90 mm. In addition, the constant filtration rate increased by 53% after the sponge city construction in the Wufeng River basin. While the rates in the road and municipal land increased by 20 times. The increasing constant filtration rater would lead the ratio of surface net precipitation to the total precipitation to decrease. In contrast, it would lead the ratio of underground net precipitation to total precipitation to increase. And the decreasing ratio of surface net precipitation triggered the overland excess infiltration weaken, followed by the decrease in the overland runoff generation capacity combined with decreasing runoff coefficient. Conversely, the increasing ratio of underground net precipitation triggered the underground runoff generation accelerated. But it could offset the influence of the decrease in the runoff coefficient to make underground runoff coefficient decrease and runoff generation capacity weaken. Moreover, the surface roughness increased by 27% after the sponge city construction in the Wufeng River basin and by 2 times for the permeable pavements. And the parameter K in the Muskingum method decreased by 91% in the same scenario. Those changes worked together to lead the unit discharge of the overland flow decrease by 37%, leading the runoff concentration slow. Lastly, the sponge facilities were characterized by source reduction, runoff concentration block, and point flood and waterlogging regulation and storage function. It distributed in the basin as the reservoir to enhance the runoff control capacity of the basin, weak the overland excess filtration runoff, and block the runoff concentration. But it strengthened the underground runoff generation.
(4) The response of the sponge city construction on the flood was evaluated, followed by its sensibility to runoff generation and concentration coefficients. Then the peak discharge and total volume of the flood were assessed to decrease by 38% and 14% after the sponge city construction for the storm of 20-30 mm, 30-40 mm, 40-90 mm, and above 90 mm, respectively. And the flood calming was observed in the same period. The sensitivity analysis of the peak discharge and total volume of the flood were implemented for CN, constant filtration rate, and surface roughness at 5 levels. Which was formulated by the empirical orthogonal function. It indicated that the sponge city construction made the decrease in CN and increase in the constant filtration rate, and surface roughness, which was the reason of the decrease in the peak discharge and total volume of the flood. Additionally, the peak discharge and total volume were detected to be more sensitive to CN. Accordingly, the decrease in the peak discharge and total volume was mainly attributed to the decrease in CN.
(5) The response of the sponge city construction on the waterlogging was evaluated. And the relationship was built between responses of the sponge city construction in the waterlogging and runoff generation and concentration. The corresponding reason was revealed. Specially, the temporally mean and maximum depth of the waterlogging were assessed to decrease by 38% and 47% respectively, in addition to the decrease in waterlogging area by 51% and waterlogging duration by 62%. In addition, the source reduction function of the sponge facilities led to the decrease in the waterlogging depth, area, and duration. In contrast, the runoff concentration block function of the sponge facilities led to the increase in the waterlogging depth, area, and duration. And the observed decrease in the waterlogging depth, area, and duration was attributed to more significant influence of the source reduction function of the sponge facilities.
|MOST Discipline Catalogue||农学::农业资源与环境|
|白桦. 海绵城市防洪减涝效应评价模型及其应用[D]. 北京. 中国科学院大学,2020.|
|Files in This Item:|
|海绵城市防洪减涝效应评价模型及其应用.p（2744KB）||学位论文||开放获取||CC BY-NC-SA||Application Full Text|
|Recommend this item|
|Export to Endnote|
|Similar articles in Google Scholar|
|Similar articles in Baidu academic|
|Similar articles in Bing Scholar|
Items in the repository are protected by copyright, with all rights reserved, unless otherwise indicated.