Amid growing typhoon risks driven by climate change with projected shifts in precipitation intensity and temperature patterns,Taiwan faces increasing challenges in flood risk.In response,this study proposes a geograph...Amid growing typhoon risks driven by climate change with projected shifts in precipitation intensity and temperature patterns,Taiwan faces increasing challenges in flood risk.In response,this study proposes a geographic information system(GIS)-based artificial intelligence(AI)model to assess flood susceptibility in Keelung City,integrating geospatial and hydrometeorological data collected during Typhoon Krathon(2024).The model employs the random forest(RF)algorithm,using seven environmental variables excluding average elevation,slope,topographic wetness index(TWI),frequency of cumulative rainfall threshold exceedance,normalized difference vegetation index(NDVI),flow accumulation,and drainage density,with the number of flood events per unit area as the output.The RF model demonstrates high accuracy,achieving the accuracy of 97.45%.Feature importance indicates that NDVI is the most critical predictor,followed by flow accumulation,TWI,and rainfall frequency.Furthermore,under the IPCC AR5 RCP8.5 scenarios,projected 50-year return period rainfall in Keelung City increases by 42.40%-64.95%under+2℃to+4℃warming.These projections were integrated into the RF model to simulate future flood susceptibility.Results indicate two districts in the study area face the greatest increase in flood risk,emphasizing the need for targeted climate adaptation in vulnerable urban areas.展开更多
Accurate soil classification is essential for pavement design;however,the traditional American Association of State Highway and Transportation Officials(AASHTO)classification system relies on extensive laboratory test...Accurate soil classification is essential for pavement design;however,the traditional American Association of State Highway and Transportation Officials(AASHTO)classification system relies on extensive laboratory testing and subjective judgment.This study presents an artificial intelligence(AI)enhanced framework for AASHTO soil classification.A synthetic dataset of 349,015 samples was generated using parameter ranges for five AASHTO input variables to support model development.Four machine learning models were trained,analyzed,and compared where the random forest(RF)consistently achieved the highest accuracy of 100%among the four models in predicting AASHTO soil groups.Feature importance analysis indicates that percent passing the No.200 sieve is the most influential factor,and under missing input scenarios.Additionally,the models remain reliable under partial input loss,though accuracy is most sensitive to the absence of percent passing the No.200 sieve,dropping to 85.8%,while all other variables maintain accuracies of at least 93.1%.Prediction uncertainty using Monte Carlo simulations shows model performance within a 95%confidence interval.Overall,the proposed AI models can accurately and efficiently predict AASHTO soil groups using incomplete datasets for geotechnical engineering.展开更多
基金supported by the National Science and Technology Council(NSTC),Taiwan(NSTC 114-2119-M-019-003).
文摘Amid growing typhoon risks driven by climate change with projected shifts in precipitation intensity and temperature patterns,Taiwan faces increasing challenges in flood risk.In response,this study proposes a geographic information system(GIS)-based artificial intelligence(AI)model to assess flood susceptibility in Keelung City,integrating geospatial and hydrometeorological data collected during Typhoon Krathon(2024).The model employs the random forest(RF)algorithm,using seven environmental variables excluding average elevation,slope,topographic wetness index(TWI),frequency of cumulative rainfall threshold exceedance,normalized difference vegetation index(NDVI),flow accumulation,and drainage density,with the number of flood events per unit area as the output.The RF model demonstrates high accuracy,achieving the accuracy of 97.45%.Feature importance indicates that NDVI is the most critical predictor,followed by flow accumulation,TWI,and rainfall frequency.Furthermore,under the IPCC AR5 RCP8.5 scenarios,projected 50-year return period rainfall in Keelung City increases by 42.40%-64.95%under+2℃to+4℃warming.These projections were integrated into the RF model to simulate future flood susceptibility.Results indicate two districts in the study area face the greatest increase in flood risk,emphasizing the need for targeted climate adaptation in vulnerable urban areas.
文摘Accurate soil classification is essential for pavement design;however,the traditional American Association of State Highway and Transportation Officials(AASHTO)classification system relies on extensive laboratory testing and subjective judgment.This study presents an artificial intelligence(AI)enhanced framework for AASHTO soil classification.A synthetic dataset of 349,015 samples was generated using parameter ranges for five AASHTO input variables to support model development.Four machine learning models were trained,analyzed,and compared where the random forest(RF)consistently achieved the highest accuracy of 100%among the four models in predicting AASHTO soil groups.Feature importance analysis indicates that percent passing the No.200 sieve is the most influential factor,and under missing input scenarios.Additionally,the models remain reliable under partial input loss,though accuracy is most sensitive to the absence of percent passing the No.200 sieve,dropping to 85.8%,while all other variables maintain accuracies of at least 93.1%.Prediction uncertainty using Monte Carlo simulations shows model performance within a 95%confidence interval.Overall,the proposed AI models can accurately and efficiently predict AASHTO soil groups using incomplete datasets for geotechnical engineering.
