Characterization of groundwater potential zones in water-scarce hardrock regions using data driven model

Influencing Factors and Risk Assessment of Precipitation-Induced Flooding in Zhengzhou, China, Based on Random Forest and XGBoost Algorithms

“…Using a combination of ACC and AUC to measure the stability and accuracy of a model can avoid erroneous results caused by skewed data that often occur in data samples [ 14 ].…”

“…Regarding the chronological order of application scenarios, existing storm and flood hazard assessments can be divided into three categories: pre-disaster assessments [ 10 ], mid-disaster follow-up monitoring and assessments, and post-disaster real-world assessments; in terms of research scales, there are studies that consider large areas and watersheds as units for overall regional risk assessment and centralised control [ 11 ]; there are also studies that consider the differences in the basic characteristics of specific cities [ 12 , 13 ]. From the perspective of research methods, most of the existing studies are based on statistical principles combined with 3S technologies [ 14 , 15 ], including hierarchical analysis, entropy value method, logistic regression method [ 16 ], BP neural network evaluation method, intelligent algorithms combined with RS-GIS technology [ 17 , 18 ], and hydrodynamic models [ 19 , 20 ], etc. ; from the selected index system, the 10 index factors of topographic index, water system index, number of consecutive rainstorm days, 1 h rainfall, 24 h cumulative rainfall, land-use type, vegetation cover, population density, old and young population ratio, and GDP were used with the highest frequency and recognition.…”

Potential hotspot modeling and monitoring of PM2.5 concentration for sustainable environmental health in Maharashtra, India

“…Different kinds of interpolation methods are available to analyze data such as Natural Neighbor Method, Nearest Neighbor Method, Kriging Method, Moving Average Method, Polynomial Regression Method, and Inverse Distance Weighted Method which are used by the different scholars in their studies (Rukundo and Cao 2012 ; Apak and Atay 2013 ; Ma et al 2019 ; Das et al 2021 ; Ruidas et al 2021 , 2022a , b ). Musikavong and Gheewala ( 2017 ) have proposed that the Kriging method can be a more efficient spatial analysis tool at the national level; whereas the method is appropriate and mostly used spatial analysis tool at the regional level (Zhang and Tripathi 2018 ).…”

Strict lockdown measures reduced PM2.5 concentrations during the COVID-19 pandemic in Kolkata, India

“…There are several interpolation methods are available to perform the spatial distribution including kriging, IDW, spline, natural neighbor, moving average, polynomial regression, etc. But IDW is preferable for the small geographical area (Basu 2015 ) and is also used by several researchers in their work (Apak and Atay 2013 ; Pal et al 2022b ; Ruidas et al 2021 , 2022 ). According to (Zhang and Tripathi 2018 ), the IDW method is the most suitable tool for regional spatial distribution analysis.…”

“…Kumar et al ( 2020 ) also showed a significant number of European countries were facing extremely worse conditions than all the countries of the World. In a global scenario, India became a vulnerable place for her huge population and population density compared to other countries and the fatal aftermath of COVID-19 creates this unprecedented massacre worldwide (Bera et al 2021a ; Ruidas and Pal 2022 ). The rapid spreading nature of this fatal disease pushed the Indian government to take immediate necessary steps to prevent the deadly situation.…”