| Urban hydrological changes caused by rapid urbanization and climate change have increased the risk of urban flooding.Low Impact Development(LID)can increase the infiltration of rainwater and reduce the risk of flooding.Therefore,LID facilities have played an important role in stormwater management and urban planning.The first step in LID construction is to understand the composition of urban underlying surface.Highresolution remote sensing images are widely used for land use information extraction because of their clear shape,rich texture and other characteristics.On the basis of accurate extraction of underlying surface,the hydrological and water quality regulation effect of low-impact development is simulated and analyzed,and different construction schemes are comprehensively evaluated which can further improve the scientific nature of sponge city planning and construction.In this paper,an area of Qingdao is taken as the research area,and the object-oriented image analysis and GF-2 high-resolution remote sensing image data are applied to extract the underlying surface type information in the study area.Storm Water Management Model(SWMM)was used to build a rainwater runoff model,and different LID combination schemes were simulated to evaluate their regulatory effects on surface runoff and non-point source pollution.An evaluation index system was established from three aspects of environmental,economic,and social and a comprehensive evaluation of LID schemes was carried out based on entropy weight-TOPSIS method.The research results provide scientific support for the design and application of LID scheme in sponge city construction in the study area,and also provide reference for sponge city construction in other regions.The main research content and conclusions of the paper are as follows:(1)Based on the object-oriented image analysis method,GF-2 satellite remote sensing image with high spatial resolution was used as the experimental data to extract the underlying surface type.The images were pre-processed by orthorectification,image fusion,atmospheric correction,and clipping,and the study area was determined to be predominantly composed of three types of land cover.Multi-scale segmentation was then performed using eCognition software,and the type and area od the underlying surface of the study area wereextracted according to the characteristics of the local objects.The accuracy of the classification results was evaluated using the Error Matrix based on TTA mask evaluation tool.The result showed that the road area was11.27 km2,the building roof area was 6.19 km2,and the green area was 3.26 km2.The overall classification accuracy was found to be 0.93 with a Kappa coefficient of 0.88,indicating a high degree of accuracy in the classification of land cover types in the study area.(2)According to the underlying surface information extracted above,the rainwater management model of the study area was constructed using SWMM.The model simulated the surface runoff and non-point source pollution of the current land cover types in the study area under the recurrence period P=3a,P=5a,P=10a,P=20a,P=50a.The result showed that both the total amount of runoff and total amount of pollutants increased with increasing rainfall intensity.The concentrations of SS,COD,and TN at the drainage outlets showed an initial increase followed by a decrease over time as rainfall continued.In addition,with the rainfall return period increased,the peak concentrations of pollutants decreased and the time to reach the peak concentration was 2-8 min earlier.(3)Taking into account the land cover types and terrain characteristics of the study area,thress facilities such as,sunken green space,permeable pavement and green roof,were selected to design six LID schemes with different area ratios.The runoff and non-point source pollution control effects of each scheme were simulated.The result showed that in terms of runoff control,the storage effect of the six schemes on runoff regulation was ranked as follows:Scheme 6>Scheme 5>Scheme 4>Scheme 3>Scheme 2>Scheme 1.The control effect of scheme is better at medium and low rainfall intensities than at high rainfall intensities.The reduction effect of all schemes on runoff was more than 30%,and the highest reduction effect was 48.3 8%.The reduction effect on peak flow was more than 12%,and the highest reduction effect was 28.75%.In terms of pollutant control,the purification and removal effect of six schemes on pollutants was as follows:Scheme 6>Scheme 5>Scheme 4>Scheme 3>Scheme 2>Scheme 1.In summary,the larger the coverage area of LID facilities,the better the control effect on surface runoff and non-point source pollution,but the relationship is non-linear.(4)From the perspectives of environmental benefits,economic benefits,and social benefits,seven evaluation indicators:runoff reduction rate,peak flow reduction rate,SS load reduction rate,COD load reduction rate,TN load reduction rate,life cycle cost,and carbon emissions were selected to construct a comprehensive evaluation index system.The entropy weight-TOPSIS method was used for comprehensive evaluation and optimization of six LID combination schemes.The results showed that the comprehensive benefits of different schemes was ranked as follows:Scheme 5>Scheme 6>Scheme 4>Scheme 2>Scheme 3>Scheme 1,indicating that scheme 5 had the highest comprehensive benefits and was the best scheme for LID construction in the study area. |
