Automated classification of building damage in remote sensing images enables the rapid and spatially extensive assessment of the impact of natural hazards, thus speeding up emergency response efforts. Convolutional neural networks (CNNs) can reach good performance on such a task in experimental settings. How CNNs perform when applied under operational emergency conditions, with unseen data and time constraints, is not well studied. This study focuses on the applicability of a CNN-based model in such scenarios. We performed experiments on 13 disasters that differ in natural hazard type, geographical location, and image parameters. The types of natural hazards were hurricanes, tornadoes, floods, tsunamis, and volcanic eruptions, which struck across North America, Central America, and Asia. We used 175,289 buildings from the xBD dataset, which contains human-annotated multiclass damage labels on high-resolution satellite imagery with red, green, and blue (RGB) bands. First, our experiments showed that the performance in terms of area under the curve does not correlate with the type of natural hazard, geographical region, and satellite parameters such as the off-nadir angle. Second, while performance differed highly between occurrences of disasters, our model still reached a high level of performance without using any labeled data of the test disaster during training. This provides the first evidence that such a model can be effectively applied under operational conditions, where labeled damage data of the disaster cannot be available timely and thus model (re-)training is not an option.

Geodata, geographical information science (GISc), and GeoAI (geo-intelligence workflows) play an increasingly important role in predictive disaster risk reduction and management (DRRM), aiding decision-makers in determining where and when to allocate resources. There have been discussions on the ethical pitfalls of these predictive systems in the context of DRRM because of the documented cases of biases in AI systems in other socio-technical systems. However, none of the discussions expound on how to audit geo-intelligence workflows for biases from data collection, processing, and model development. This paper considers a case study that uses AI to characterize housing stock vulnerability to flooding in Karonga district, Malawi. We use Friedman and Nissenbaum’s definition and categorization of biases that emphasize biases as a negative and undesirable outcome. We limit the scope of the audit to biases that affect the visibility of different housing typologies in the workflow. The results show how AI introduces and amplifies these biases against houses of certain materials. Hence, a group within the population in the area living in these houses would potentially miss out on DRRM interventions. Based on this example, we urge the community of researchers and practitioners to normalize the auditing of geo-intelligence workflows to prevent information disasters from biases.

Epidemics are among the most costly and destructive natural hazards globally. To reduce the impacts of infectious disease outbreaks, the development of a risk index for infectious diseases can be effective, by shifting infectious disease control from emergency response to early detection and prevention.In this study, we introduce a methodology to construct and validate an epidemic risk index using only open data, with a specific focus on scalability. The external validation of our risk index makes use of distance sampling to correct for underreporting of infections, which is often a major source of biases, based on geographical accessibility to health facilities. We apply this methodology to assess the risk of dengue in the Philippines.The results show that the computed dengue risk correlates well with standard epidemiological metrics, i.e. dengue incidence (p = 0.002). Here, dengue risk constitutes of the two dimensions susceptibility and exposure. Susceptibility was particularly associated with dengue incidence (p = 0.048) and dengue case fatality rate (CFR) (p = 0.029). Exposure had lower correlations to dengue incidence (p = 0.193) and CFR (p = 0.162). Highest risk indices were seen in the south of the country, mainly among regions with relatively high susceptibility to dengue outbreaks.Our findings reflect that the modelled epidemic risk index is a strong indication of sub-national dengue disease patterns and has therefore proven suitability for disease risk assessments in the absence of timely epidemiological data. The presented methodology enables the construction of a practical, evidence-based tool to support public health and humanitarian decision-making processes with simple, understandable metrics. The index overcomes the main limitations of existing indices in terms of construction and actionability.

Anisotropic flow is a key observable to characterise the system created in heavy-ion collisions, as it is sensitive to the system's initial state, transport properties, the equation of state and freeze-out conditions. In these proceedings we present the anisotropic flow coefficients of inclusive charged particles in Pb–Pb collisions at $\sqrt{s_{\mathrm{NN}}}=2.76\text{and}5.02\text{TeV}$, and in Xe–Xe collisions at $\sqrt{s_{\mathrm{NN}}}=5.44\text{TeV}$. The results are reported for a wide range of particle transverse momentum within the pseudo-rapidity range |η| < 0.8 at different collision centralities. The energy and system dependence are found to place strong constraints on the temperature dependence of η/s and the modeling of the initial state, respectively. We also present detailed studies of flow fluctuations in heavy-ion collisions, in order to precisely characterise the underlying flow probability density function. We find evidence of non-Bessel-Gaussian fluctuations and discuss the origin of this observation.

To accurately identify the most vulnerable areas to floods, physical (e.g., building material) and social (e.g., education, health, income of households) housing stock information is required. However, in developing countries, this information is often unreliable, unavailable or inaccessible, and manual data collection is time-consuming. This can lead to difficulties for humanitarians or policymakers in implementing appropriate disaster risk reduction and response interventions. Therefore, there is a need for the development of alternative approaches to data collection and analysis. An alternative approach to on-site vulnerability assessment is to extract physical vulnerability characteristics, such as land use type or rooftop material, from satellite or Unmanned Aerial Vehicle (UAV) imagery. However, other social or physical vulnerability information on the household level can often not be extracted from only the remote sensing data. This research develops an approach for integrating multiple data sources into a Geographic Information System to improve the completeness of data on different vulnerability indicators. This approach is applied on the housing stock of the Karonga district in Malawi. An Object-Based Image Analysis on UAV imagery is combined with a machine learning analysis of Mapillary data to enable remote identification of both rooftop ànd wall material. Depth-damage curves were created to describe the impact on the housing stock for different categories of physical vulnerability (such as building material) and levels of inundation. Moreover, local survey data is used for the creation of a social vulnerability index. Combined, the datasets represent the spatial distribution of housing stock vulnerability for multiple flood scenarios. This approach is useful in situations where proactive risk analyses must be carried out or where local-scale interventions, such as building strengthening- or flood awareness projects, have to be implemented. Finally, we give recommendations for scaling the methodology to areas where only lower resolution data is available.

Structured datasets of loss data, i.e. data on the impact of past natural disasters, are of paramount importance for informing disaster preparedness programs and forecasting the impact of future disasters. Most of existing initiatives aim at manually building such datasets from information of goverments, humanitarian agencies and researchers. Unfortunately, the quality and completeness of such information is often insufficient, especially for small disasters and/or in areas where these organisations are not active. More often, it's local and national newspapers that report on small disasters. In this contribution, we present a series of algorithms to automatically extract structured loss data from online newspapers, even small ones that are not captured by common news aggregator (e.g. Google News). The algorithms are validated both in terms of accuracy of extraction and consistency with existing datasets; we argue that they provide a valuable tool to collect loss data in data-poor regions.