98%
921
2 minutes
20
Machine Learning (ML) models have become a pivotal tool in the scientific community, successfully addressing complex problems across various domains, including flood risk management. Despite these advancements, traditional data-driven models often struggle when training data is scarce and primarily rely on correlation rather than causal relationships, making them vulnerable to shifts in data distribution. This paper introduces a Causally Informed Neural Network (CINN) framework that integrates causal prior knowledge as an inductive bias to improve flood damage predictions for residential properties and address these limitations. The proposed approach enhances model adaptability to unseen data distributions-an essential requirement for flood damage modeling. First, Deep End-to-End Causal Inference (DECI) is used to discover causal relationships and estimate their average treatment effects. These causal insights are then embedded into the neural network through causal weight initialization and causal regularization. The framework is validated using an enhanced National Flood Insurance Program (NFIP) claims dataset from Hurricane Katrina, and its performance is benchmarked against six widely used ML models from previous studies. Results show that the discovered causal relationships align with domain knowledge, reinforcing the approach's credibility. The proposed CINN model achieves an average 22 % error reduction compared to traditional ML models, demonstrating its superior robustness and predictive accuracy. Additionally, a feature attribution experiment confirms that the model's decision-making process is consistent with the identified causal relationships, increasing interpretability and trust in its predictions. These findings highlight the potential of integrating causality into ML-based flood damage modeling, paving the way for more resilient and generalizable disaster risk assessment frameworks.
Download full-text PDF |
Source |
---|---|
http://dx.doi.org/10.1016/j.scitotenv.2025.180121 | DOI Listing |
PLoS One
September 2025
Daqing Yongzhu Petroleum Technology Development Co Ltd., Daqing, China.
Background: Strongly water-sensitive reservoirs with high clay content face challenges in conventional development due to clay swelling and impeded seepage. CO2 injection shows potential for enhanced oil recovery (EOR) and carbon sequestration; however, the role of clay minerals in regulating CO2-induced asphaltene deposition and sequestration remains unclear.
Methodology: We conducted experiments on clay-oil interactions, nuclear magnetic resonance (NMR), measurements of crude oil properties, and long core water flooding tests to evaluate deposition, reservoir damage, and CO2 sequestration.
Biology (Basel)
July 2025
Department of Biochemistry and Microbiology, Rutgers, The State University of New Jersey, New Brunswick, NJ 08901, USA.
Molds readily grow on wet books, documents, and other library materials where they ruin them chemically, mechanically, and aesthetically. Poor maintenance of libraries, failures of Heating, Ventilation, and Air Conditioning (HVAC) systems, roof leaks, and storm damage leading to flooding can all result in accelerated fungal growth. Moreover, when fungal spores are present at high concentrations in the air, they can be linked to severe respiratory conditions and possibly to other adverse health effects in humans.
View Article and Find Full Text PDFSci Rep
September 2025
Department of Civil Engineering, Pulchowk Campus, Institute of Engineering, Tribhuvan University, Lalitpur, Nepal.
Floods are among the most damaging natural disasters, posing significant threats to socio-economic stability and environmental sustainability. This study addresses an important research gap by evaluating flood susceptibility in a small watershed (< 500 km), where no detailed susceptibility mapping has been conducted before. Flood susceptibility in the Triyuga Watershed, Nepal, was evaluated using three statistical models: Frequency Ratio (FR), Logistic Regression (LR), and Weight of Evidence (WoE), and the distinct hydrological behaviours of small watersheds were highlighted.
View Article and Find Full Text PDFCamb Prism Coast Futur
February 2025
Institute for Environmental Studies, Vrije Universiteit Amsterdam, Amsterdam, The Netherlands.
The record storm surge of October 2023, which hit the southwestern German Baltic Sea, not only resulted in significant damages to coastal communities and infrastructure but also demonstrated that the region was prepared and able to avoid loss of lives and other catastrophic impacts. Numerical modelling has been a key tool utilised for providing information to support coastal flood management, at different levels of planning, for such events. Based on recent research conducted in the Baltic coast region as well as on empirical evidence acquired during the event, we present an operational scheme that utilises modelling tools and frameworks for supporting coastal flood management in the region.
View Article and Find Full Text PDFInt J Mol Sci
August 2025
Faculty of Environment and Information Sciences, Fukui University of Technology, Fukui 910-8505, Japan.
Plants are constantly exposed to various environmental challenges, such as drought, flooding, heavy metal toxicity, and pathogen attacks. To cope with these stresses, they employ several adaptive strategies. This review highlights the potential of plant-derived smoke (PDS) solution as a natural biostimulant for improving plant health and resilience, contributing to both crop productivity and ecological restoration under abiotic and biotic stress conditions.
View Article and Find Full Text PDF