Explore the vast range of titles available.

Deep learning techniques have been increasingly used in flood management to overcome the limitations of accurate, yet slow, numerical models and to improve the results of traditional methods for flood mapping. In this paper, we review 58 recent publications to outline the state of the art of the field, identify knowledge gaps, and propose future research directions. The review focuses on the type of deep learning models used for various flood mapping applications, the flood types considered, the spatial scale of the studied events, and the data used for model development. The results show that models based on convolutional layers are usually more accurate, as they leverage inductive biases to better process the spatial characteristics of the flooding events. Models based on fully connected layers, instead, provide accurate results when coupled with other statistical models. Deep learning models showed increased accuracy when compared to traditional approaches and increased speed when compared to numerical methods. While there exist several applications in flood susceptibility, inundation, and hazard mapping, more work is needed to understand how deep learning can assist in real-time flood warning during an emergency and how it can be employed to estimate flood risk. A major challenge lies in developing deep learning models that can generalize to unseen case studies. Furthermore, all reviewed models and their outputs are deterministic, with limited considerations for uncertainties in outcomes and probabilistic predictions. The authors argue that these identified gaps can be addressed by exploiting recent fundamental advancements in deep learning or by taking inspiration from developments in other applied areas. Models based on graph neural networks and neural operators can work with arbitrarily structured data and thus should be capable of generalizing across different case studies and could account for complex interactions with the natural and built environment. Physics-based deep learning can be used to preserve the underlying physical equations resulting in more reliable speed-up alternatives for numerical models. Similarly, probabilistic models can be built by resorting to deep Gaussian processes or Bayesian neural networks.

A brief overview of the methods of a tsunami early warning in the Kuril Islands, which turned out to be ineffective during recent events, is presented. A hydrophysical method for short-term tsunami forecasting based on information about a tsunami in the ocean, used in the United States, and an express method, also using information about a tsunami in the ocean, are briefly described. The results of the retrospective forecast of the tsunami that occured on March 11, 2011, by the express method are presented.

Pysteps is an open-source and community-driven Python library for probabilistic precipitation nowcasting, that is, very-short-range forecasting (0–6 h). The aim of pysteps is to serve two different needs. The first is to provide a modular and well-documented framework for researchers interested in developing new methods for nowcasting and stochastic space–time simulation of precipitation. The second aim is to offer a highly configurable and easily accessible platform for practitioners ranging from weather forecasters to hydrologists. In this sense, pysteps has the potential to become an important component for integrated early warning systems for severe weather.The pysteps library supports various input/output file formats and implements several optical flow methods as well as advanced stochastic generators to produce ensemble nowcasts. In addition, it includes tools for visualizing and post-processing the nowcasts and methods for deterministic, probabilistic and neighborhood forecast verification. The pysteps library is described and its potential is demonstrated using radar composite images from Finland, Switzerland, the United States and Australia. Finally, scientific experiments are carried out to help the reader to understand the pysteps framework and sensitivity to model parameters.

Two successive landslides within a month started in October 11, 2018, and dammed twice the Jinsha River at the border between Sichuan Province and Tibet in China. Both events had potential to cause catastrophic flooding that would have disrupted lives of millions and induced significant economic losses. Fortunately, prompt action by local authorities supported by the deployment of a real-time landslide early warning system allowed for quick and safe construction of a spillway to drain the dammed lake. It averted the worst scenario without loss of life and property at least one order of magnitude less to what would have been observed without quick intervention. Particularly, the early warning system was able to predict the second large-scale slope failure 24 h in advance, along with minor rock falls during the spillway construction, avoiding false alerts. This paper presents the main characteristics of both slope collapses and damming processes, and introduces the successful landslide early warning system. Furthermore, we found that the slope endured cumulative creeping displacements of > 40 m in the past decade before the first event. Twenty-five meter displacement occurred in the year immediately before. The deformation was measured by the visual interpretation of multitemporal satellite images, which agrees with the interferometry synthetic aperture radar (InSAR) measurement. If these had been done before the emergency, economic losses could have been reduced further. Therefore, our findings strengthen the case for the deployment of systematic monitoring of potential landslide sites by integrating earth observation methods (i.e., multitemporal satellite or UAV images) and in situ monitoring system as a way to reduce risk. It is expected that this success story can be replicated worldwide, contributing to make our society more resilient to landslide events.

Recent years include the world’s hottest year, while they have been marked mainly, besides the COVID-19 pandemic, by climate-related disasters, based on data collected by the Emergency Events Database (EM-DAT). Besides the human losses, disasters cause significant and often catastrophic socioeconomic impacts, including economic losses. Recent developments in artificial intelligence (AI) and especially in machine learning (ML) and deep learning (DL) have been used to better cope with the severe and often catastrophic impacts of disasters. This paper aims to provide an overview of the research studies, presented since 2017, focusing on ML and DL developed methods for disaster management. In particular, focus has been given on studies in the areas of disaster and hazard prediction, risk and vulnerability assessment, disaster detection, early warning systems, disaster monitoring, damage assessment and post-disaster response as well as cases studies. Furthermore, some recently developed ML and DL applications for disaster management have been analyzed. A discussion of the findings is provided as well as directions for further research.

Focus on behaviors, not on fetishizing social media tools Crisis informatics is a multidisciplinary field combining computing and social science knowledge of disasters; its central tenet is that people use personal information and communication technology to respond to disaster in creative ways to cope with uncertainty. We study and develop computational support for collection and sociobehavioral analysis of online participation (i.e., tweets and Facebook posts) to address challenges in disaster warning, response, and recovery. Because such data are rarely tidy, we offer lessons—learned the hard way, as we have made every mistake described below—with respect to the opportunities and limitations of social media research on crisis events.

Tsunami warning centres face the challenging task of rapidly forecasting tsunami threat immediately after an earthquake, when there is high uncertainty due to data deficiency. Here we introduce Probabilistic Tsunami Forecasting (PTF) for tsunami early warning. PTF explicitly treats data- and forecast-uncertainties, enabling alert level definitions according to any predefined level of conservatism, which is connected to the average balance of missed-vs-false-alarms. Impact forecasts and resulting recommendations become progressively less uncertain as new data become available. Here we report an implementation for near-source early warning and test it systematically by hindcasting the great 2010 M8.8 Maule (Chile) and the well-studied 2003 M6.8 Zemmouri-Boumerdes (Algeria) tsunamis, as well as all the Mediterranean earthquakes that triggered alert messages at the Italian Tsunami Warning Centre since its inception in 2015, demonstrating forecasting accuracy over a wide range of magnitudes and earthquake types. Probabilistic tsunami forecasting (PTF) defines an approach to tsunami early warning based on uncertainty quantification, enhancing forecast accuracy and enabling rational decision making. PTF is here developed for near-source tsunami warning, and tested in hindcasting mode over a wide range of past earthquakes.

In hospitalized patients, the risk of sepsis-related mortality can be assessed using the quick Sepsis-related Organ Failure Assessment (qSOFA). Currently, different tools that predict deterioration such as the National Early Warning Score (NEWS) have been introduced in clinical practice in Emergency Departments (ED) worldwide. It remains ambiguous which screening tool for mortality at the ED is best. The objective of this study was to evaluate the predictive performance for mortality of two sepsis-based scores (i.e. qSOFA and Systemic Inflammatory Response Syndrome (SIRS)-criteria) compared to the more general NEWS score, in patients with suspected infection directly at presentation to the ED. We performed a retrospective cohort study. Patients who presented to the ED between June 2012 and May 2016 with suspected sepsis in a large tertiary care center were included. Suspected sepsis was defined as initiation of intravenous antibiotics and/or collection of any culture in the ED. Outcome was defined as 10-day and 30-day mortality after ED presentation. Predictive performance was expressed as discrimination (AUC) and calibration using Hosmer-Lemeshow goodness-of-fit test. Subsequently, sensitivity, and specificity were calculated. In total 8,204 patients were included of whom 286 (3.5%) died within ten days and 490 (6.0%) within 30 days after presentation. NEWS had the best performance, followed by qSOFA and SIRS (10-day AUC: 0.837, 0.744, 0.646, 30-day AUC: 0.779, 0.697, 0.631). qSOFA (≥2) lacked a high sensitivity versus SIRS (≥2) and NEWS (≥7) (28.5%, 77.2%, 68.0%), whilst entailing highest specificity versus NEWS and SIRS (93.7%, 66.5%, 37.6%). NEWS is more accurate in predicting 10- and 30-day mortality than qSOFA and SIRS in patients presenting to the ED with suspected sepsis.

The utilization of deep mineral resources is severely restricted by dynamic disasters, such as rock burst, caused by mining disturbances. This study is aimed at investigating the instability failure mechanism and precursory response of coal specimen, with the aid of acoustic emission (AE) monitoring, under triaxial loading and unloading disturbance. The AE time-varying characteristics were analyzed, the crack propagation mode was explored, and the statistical b value and βt value (damage parameters) of AE events were introduced to evaluate the damage state of the loaded coal under disturbance. The results show that: the dynamic disturbance loads will accelerate the deformation process of coal specimen under high static confining stress, resulting in a positive relation between the AE characteristic parameters (AE counts and energy) and the increasing disturbance amplitude. The proportion of low-frequency and high-amplitude AE signals increases with the disturbance amplitude, indicating the disturbance with large amplitude will promote the formation of large-magnitude cracking events. Those stress-induced cracks, judged from the RA-AF value of AE signals, are mainly tensile mode (i.e., mode-I) under the applied disturbance load, but shear cracks are the critical factor affecting the instability and failure of the loaded coal specimen. Both the b value and βt value can reflect the time-varying evolution process of internal damage for the loaded coal specimen. The sudden decreasing of the b value together with the exponential increasing of the βt value can be regarded as a short-impending precursor indicator for the coal instability and failure. The results of this study will provide some experiment basic and theoretical guidance for dynamic disaster monitoring and early-warning during the deep coal mining.HighlightsThe instability failure mechanism and the precursory response of coal specimens under triaxial loading and unloading disturbance were investigated.By analyzing the AE characteristic parameters, the internal damage evolution processes, crack propagation mode and the mechanisms of disturbance were studied.The damage accumulation process and precursor warning information of coal instability failure are extracted by the b and βt values.