Recent unprecedented events have highlighted that the existing approach to managing flood risk is inadequate for complex urban systems because of its overreliance on simplistic methods at coarse‐resolution large scales, lack of model physicality using loose hydrologic‐hydraulic coupling, and absence of urban water infrastructure at large scales. Distributed models are a potential alternative as they can capture the complex nature of these events through simultaneous tracking of hydrologic and hydrodynamic processes. However, their application to large‐scale flood mapping and forecasting remains challenging without compromising on spatiotemporal resolution, spatial scale, model accuracy, and local‐scale hydrodynamics. Therefore, it is essential to develop techniques that can address these issues in urban systems while maximizing computational efficiency and maintaining accuracy at large scales. This study presents a physically based but computationally efficient approach for large‐scale (area > 10³ km²) flood modeling of extreme events using a distributed model called Interconnected Channel and Pond Routing. The performance of the proposed approach is compared with a hyperresolution‐fixed‐mesh model at 60‐m resolution. Application of the proposed approach reduces the number of computational elements by 80% and the simulation time for Hurricane Harvey by approximately 4.5 times when compared to the fixed‐resolution model. The results show that the proposed approach can simulate the flood stages and depths across multiple gages with a high accuracy (R² > 0.8). Comparison with Federal Emergency Management Agency building damage assessment data shows a correlation greater than 95% in predicting spatially distributed flooded locations. Finally, the proposed approach can estimate flood stages directly from rainfall for ungaged streams.

Automation in structural health monitoring has generated a lot of interest in recent years, especially with the introduction of cheap digital cameras. This paper presents fuzzy logic and artificial neural network based models for accurate crack detection on concrete. Features are extracted from digital images of concrete surfaces using image processing which incorporates the edge detection technique. The properties of extracted features are fed into the models for detecting cracks. Two kinds of approaches have been implemented in this study: the image approach which classifies an image as a whole, and the object approach which classifies each component or object in an image into cracks and noise. The models have been tested on 205 images and evaluated on the basis of five measures of performance.

The effect of climate change on precipitation intensity is well documented. However, findings regarding changes in spatial extent of extreme precipitation events are still ambiguous as previous studies focused on particular regions and time domains. This study addresses this ambiguity by investigating the pattern of changes in the spatial extent of short duration extreme precipitation events globally. A grid‐based indicator termed Spatial‐Homogeneity is proposed and used to assess the changes of spatial extent in Global Precipitation Measurement records. This study shows that (a) rising temperature causes significant shrinking of precipitation extent in tropics, but an expansion of precipitation extent in arid regions, (b) storms with higher precipitation intensity show a faster decrease in spatial extent, and (c) larger spatial extent storms are associated with higher total precipitable water. Results imply that in a warming climate, tropics may experience severe floods as storms may become more intense and spatially concentrated.

Estimating uncertainty in flood model predictions is important for many applications, including risk assessment and flood forecasting. We focus on uncertainty in physics‐based urban flooding models. We consider the effects of the model's complexity and uncertainty in key input parameters. The effect of rainfall intensity on the uncertainty in water depth predictions is also studied. As a test study, we choose the Interconnected Channel and Pond Routing (ICPR) model of a part of the city of Minneapolis. The uncertainty in the ICPR model's predictions of the floodwater depth is quantified in terms of the ensemble variance using the multilevel Monte Carlo (MC) simulation method. Our results show that uncertainties in the studied domain are highly localized. Model simplifications, such as disregarding the groundwater flow, lead to overly confident predictions, that is, predictions that are both less accurate and uncertain than those of the more complex model. We find that for the same number of uncertain parameters, increasing the model resolution reduces uncertainty in the model predictions (and increases the MC method's computational cost). We employ the multilevel MC method to reduce the cost of estimating uncertainty in a high‐resolution ICPR model. Finally, we use the ensemble estimates of the mean and covariance of the flood depth for real‐time flood depth forecasting using the physics‐informed Gaussian process regression method. We show that even with few measurements, the proposed framework results in a more accurate forecast than that provided by the mean prediction of the ICPR model.