Study On Urban Flood Risk Multi-scenario Simulation Based On SD-PLUS Model And MaxEnt Model

In the process of urbanization,the unreasonable use of land has seriously disrupted the natural rainwater cycle in cities,leading to frequent urban flooding.This has had a significant impact on various human activities based on land and the sustainable development of cities.Most previous studies on the assessment of urban flooding risks under land use and land cover(LULC)changes were based on historical or current data,and research on dynamic assessments of urban flooding risks combined with future urban development is relatively scarce.Therefore,this study selected the Guangdong-Hong Kong-Macao Greater Bay Area(GBA)as the study area to explore the response relationship between LULC changes under multiple scenarios and urban flood disasters,and explain their universal laws.Based on this,scientific measures are proposed for flood control and disaster reduction planning in the GBA to promote its sustainable development.Firstly,this article selects SSP126,SSP245,and SSP585 as the development scenarios for future cities from the combination of Shared Socioeconomic Pathways(SSP)and Representative Concentration Pathways(RCP),and predicts the LULC demand under different development scenarios in a System Dynamic(SD)model.Secondly,by integrating the Patch-generating Land Use Simulation Model(PLUS),which incorporates the Random Forest algorithm to identify the causes of various types of land use change,the LULC spatial patterns of the GBA under different development scenarios and time periods are generated.Based on this,the Maximum Entropy(Max Ent)model is used to simulate urban flood risk maps under different scenarios,revealing the changing patterns of potential urban flood risk under different LULC changes in the future,and further analyzing the response relationship between urban flood drivers and urban flood risk.The main conclusions are as follows:1)The SD model,based on dominant factors such as population,economy,and climate,is a method to deeply understand the mechanism of LULC changes and accurately simulate the demand for LULC in the GBA in the future.By analyzing the simulation results of different development scenarios in the GBA,it can be concluded that under the SSP126 scenario,the future socio-economic development in the GBA relies less on land resources.Specifically,there will be relatively less demand for construction land,and the area of forest land will show a trend of first decreasing and then increasing,but it will still decrease by about 1000 km~2in 2050.Therefore,protection and restoration measures for various ecological land types still need to be strengthened.Under the SSP245 scenario,the scale of land resources required for economic and social development will be between SSP126 and SSP245,and the demand for construction land will continue to increase in the future,resulting in significant decreases in the areas of forest,grassland,and arable land compared to the SSP126 scenario.Under the SSP585 scenario,rapid urbanization and high-speed socio-economic development will result in a significant expansion of urban construction land,and the areas of arable land,grassland,and forest land will gradually shrink over time.2)Based on the SD model’s prediction of future LULC demand in the GBA,the PLUS model effectively simulated the spatial patterns of LULC in various scenarios.The simulation results show that there are significant differences between SSP126,SSP245,and SSP585.However,overall,the construction land in all three scenarios will expand outward from the existing urban areas,and this trend will become more pronounced over time.In the SSP126 scenario,the expansion rate and trend of construction land are relatively slow.However,if no measures are taken,by 2050,one-third of the farmland and over 60%of the grassland will be occupied.In the SSP245 scenario,the range of construction land expansion will increase,and urban construction land will spread outward from the existing areas,resulting in a more significant trend of cropland,grassland,and woodland shrinking.In the SSP585 scenario,the expansion range of construction land is the largest,and a large amount of cropland,woodland,and grassland in the GBA will be occupied by 2050,resulting in severe damage to the ecological space of the cities.3)A Max Ent model was constructed using historical urban flooding points and driving factors to simulate the urban flooding risk areas in the GBA.The ROC value of the Max Ent model was 0.876,indicating good performance in simulating and predicting the urban flooding risk in the GBA.The results showed that the proportion of impervious surfaces,distance to major roads,and green space ratio were the most important variables affecting urban flooding risk.Over the next 30 years,compared with other scenarios,SSP585showed the most significant increase in potential urban flooding risk areas,with the highest number of high-risk areas.Under the SSP245 scenario,as the proportion of impervious surfaces in the city expands,the low and medium-risk urban areas gradually increase.There was no significant change in urban flooding risk areas under the SSP126 scenario.Through comparative analysis of the simulation results of various scenarios of urban flooding risk in the GBA,this study proposes relevant countermeasures and suggestions for spatial development and regulation,flood control and disaster reduction,and other aspects of the GBA in the future.The aim is to provide scientific support for improving the resilience of cities to urban flooding risk and promoting sustainable socio-economic development in the GBA.