Search

Flood events are becoming more severe, causing significant problems to human communities, including physical, psychological, and material damage. For both flood forecasting in emergency response situations and flood mapping, georeferencing and data curation are paramount in the context of prevention or preparedness. Hence, data display, data management, and articulation with numerical simulation results must occur on GIS platforms. Our research is motivated by recent advances in Web and GIS technologies, social sensing and high-performance computing, and an envisaged wider availability of remote sensing data. This paper presents and discusses an innovative Web GIS platform named "RiverCure Portal" or "RCP" for short. This platform combines observations and hydrodynamic modelling tools to support various stages of the flood risk management cycle, including operational response, emergency preparedness, and risk assessment. RCP is a multi-organization, multi-context digital platform with flexible configuration features to define and support multiple sensor types and modelling options, satisfying the various needs of different organizations and stakeholders. In addition, this paper discusses the RiverCure Approach, which encompasses the following tasks directly supported by the RCP platform: defining the context and involved geometries, associating sensors to the context, pre-processing and generating the context mesh, defining the simulation event, running the simulation event, and analyzing the results from the simulation event. Thus, the RCP streamlines and simplifies data analysis and simulation procedures to meet decision-makers' needs. The novelties discussed in this paper include the design and discussion of a Web GIS platform that allows (i) to manage flood data and results of simulations at several contextual levels by different stakeholders such as domain experts, decision-makers, researchers, or students; (ii) to process and curate sensed data obtained from physical and social sensors; and (iii) update the state and values of the parameters of simulation tools through continuous data assimilation techniques for forecasting purposes. Finally, this paper supports the explanation and discussion with a running example, "Águeda 2016 flood" event, which dataset is publicly available for further study and experimentation. [ABSTRACT FROM AUTHOR]

Geographic information has been traditionally produced by mapping agencies and corporations, using highly skilled professionals as well as expensive precision equipment and procedures, in a very costly approach. The production of land use and land cover databases is just one example of such traditional approaches. At the same time, the amount of Geographic Information created and shared by citizens through the web has been increasing exponentially during the last decade as a result of the emergence and popularization of technologies such as theWeb 2.0, cloud computing, global positioning systems (GPS), smart phones, among others. This vast amount of free geographic data might have valuable information to extract. Combining data from several initiatives might further increase the value of such data. We propose a conceptual model to integrate data from suitable user generated spatial content initiatives. A prototype to demonstrate the ability of the model to perform such integration, based on two identified use cases, was also developed. [ABSTRACT FROM AUTHOR]

This article describes a novel approach for toponym resolution with deep neural networks. The proposed approach does not involve matching references in the text against entries in a gazetteer, instead directly predicting geo-spatial coordinates. Multiple inputs are considered in the neural network architecture (e.g., the surrounding words are considered in combination with the toponym to disambiguate), using pre-trained contextual word embeddings (i.e., ELMo or BERT) as well as bi-directional Long Short-Term Memory units, which are both regularly used for modeling textual data. The intermediate representations are then used to predict a probability distribution over possible geo-spatial regions, and finally to predict the coordinates for the input toponym. The proposed model was tested on three datasets used on previous toponym resolution studies, specifically the (i) War of the Rebellion, (ii) Local–Global Lexicon, and (iii) SpatialML corpora. Moreover, we evaluated the effect of using (i) geophysical terrain properties as external information, including information on elevation or terrain development, among others, and (ii) additional data collected from Wikipedia articles, to further help with the training of the model. The obtained results show improvements using the proposed method, when compared to previous approaches, and specifically when BERT embeddings and additional data are involved. [ABSTRACT FROM AUTHOR]