In　recent　years,　with　the　global　climate　change,　intra-urban　rainstorms　are　frequent.　The　backward　means　of　urban　waterlogging　prevention　and　control　do　not　match　with　the　high-speed　urbanization　process,　bringing　serious　waterlogging　disasters　to　major　cities　in　China.　The　current　mainstream　urban　flooding　early　warning　system　is　an　integrated　system　integrating　various　information　technologies　with　urban　rainfall　and　flood　model　as　the　theoretical　basis.　However,　the　urban　rainfall　and　flooding　model　is　affected　by　the　lack　of　basic　data,　the　complicated　modeling　process　which　is　not　easy　to　implement,　and　the　poor　flexibility　in　analyzing　the　actual　urban　waterlogging　time　series　characteristics.　Therefore,　this　paper　designs　and　implements　a　waterlogging　early　warning　system　based　on　an　adaptive　urban　flooding　model　by　combining　the　respective　advantages　of　data-driven　technology　and　urban　rainfall　model　for　impervious　areas　in　cities　that　are　prone　to　flooding　during　short-duration　rainstorms.　　©　2023　IEEE.