Abstract: A method for evaluating terrain uncertainty in flood warning and forecasting is provided. The method includes: S1, acquiring three types of digital elevation model (DEM) data from a shuttle radar topography mission (SRTM), an advanced spaceborne thermal emission and reflection radiometer (ASTER), and an advanced land observing satellite (ALOS); and preprocessing the three types of DEM data; S2, optimizing urban terrain characteristics; S3, constructing a multidimensional parameter space by using Latin hypercube sampling (LHS); S4, calculating flood hydrodynamics numerical value based on multidimensional sample points; and S5, constructing a global sensitivity analysis method frame suitable for urban terrain characteristics-related factors, where a Sobol quantitative method is used in the global sensitivity analysis method frame, and the Sobol quantitative method is used to evaluate uncertainties and sensitivity characteristics of multiple factors of terrain data based on a variance decomposition theory.

Abstract: A multi-dimensional feature identification method of a disaster-causing cyclone includes: determining a disaster-causing range based on a path of the disaster-causing cyclone, dividing the disaster-causing range into multiple grid points, obtaining meteorological data including wind speeds and rainfall amounts of the grid points within the disaster-causing range; determining, based on historical typhoon data, disaster-causing thresholds including a wind speed disaster-causing threshold and a rainfall disaster-causing threshold for the meteorological data; constructing a multi-dimensional feature database for disaster-causing events; and performing multi-dimensional feature identification on the disaster-causing cyclone based on the multi-dimensional feature database for disaster-causing events.

Abstract: A high-temperature disaster forecast method based on a directed graph neural network is provided, and the method includes the following steps: S1, performing standardization processing on meteorological elements respectively to scale the meteorological elements into a same value range; S2, taking the meteorological elements as nodes in the graph, and describing relationships among the nodes by an adjacency matrix of graph; then learning node information by a stepwise learning strategy and continuously updating a state of the adjacency matrix; S3, training the directed graph neural network model after determining a loss function, obtaining a model satisfying requirements by adjusting a learning rate, an optimizer and regularization parameters as a forecast model, and saving the forecast model; and S4, inputting historical multivariable time series into the forecast model, changing an output stride according to demands, and thereby obtaining high-temperature disaster forecast for a future period of time.

Abstract: A high-temperature disaster forecast method based on a directed graph neural network is provided, and the method includes the following steps: S1, performing standardization processing on meteorological elements respectively to scale the meteorological elements into a same value range; S2, taking the meteorological elements as nodes in the graph, and describing relationships among the nodes by an adjacency matrix of graph; then learning node information by a stepwise learning strategy and continuously updating a state of the adjacency matrix; S3, training the directed graph neural network model after determining a loss function, obtaining a model satisfying requirements by adjusting a learning rate, an optimizer and regularization parameters as a forecast model, and saving the forecast model; and S4, inputting historical multivariable time series into the forecast model, changing an output stride according to demands, and thereby obtaining high-temperature disaster forecast for a future period of time.

Abstract: An evaluation method for evaluating a precipitation-induced landslide disaster loss under climate change is provided. The evaluation method belongs to the technical field of geological disaster prevention and treatment. The evaluation method uses a physical process based model, in considering of spatial heterogeneity of land-surface features of grids in the area, to obtain precipitation thresholds corresponding to the respective grids in the area having the spatial heterogeneity. Historical data and climate model data are taken in combination to select suitable climate models, and the model then is used to simulate landslide prone zones and possible influence zones caused by landslides. An influence zones simulated by the evaluation method can better match disaster loss grid data, which can solve the problem that climate change scenarios and influence of landslide are difficult to be evaluated in landslide disaster evaluation.

Abstract: An evaluation method for evaluating a precipitation-induced landslide disaster loss under climate change is provided. The evaluation method belongs to the technical field of geological disaster prevention and treatment. The evaluation method uses a physical process based model, in considering of spatial heterogeneity of land-surface features of grids in the area, to obtain precipitation thresholds corresponding to the respective grids in the area having the spatial heterogeneity. Historical data and climate model data are taken in combination to select suitable climate models, and the model then is used to simulate landslide prone zones and possible influence zones caused by landslides. An influence zones simulated by the evaluation method can better match disaster loss grid data, which can solve the problem that climate change scenarios and influence of landslide are difficult to be evaluated in landslide disaster evaluation.

Abstract: The present application provides a method and system for identifying extreme climate events. The method acquires climate index (CI) grid data of a to-be-identified region within an extreme climate time period, and gradually expands each of event centers in the to-be-identified region, until CI values of all grids adjacent to the event center are not greater than a CI threshold. The method can obtain extreme climate impacted areas of extreme climate events in the to-be-identified region, and can further obtain CI intensities of the extreme climate events by average calculation. The method can obtain three pieces of dimension information on each of the extreme climate events in the to-be-identified region, including an extreme climate impacted area, a CI intensity and a duration. Therefore, the method can identify the extreme climate events more comprehensively.

Abstract: A method for flood disaster monitoring and disaster analysis based on vision transformer is provided. It includes: step (1), constructing a bi-temporal image change detection model based on vision transformer; step (2), selecting bi-temporal remote sensing images to make flood disaster labels; and step (3), performing flood monitoring and disaster analysis according to the bi-temporal image change detection model constructed in the step (1). In combination with the bi-temporal image change detection model based on an advanced vision transformer in deep learning and radar data which is not affected by time and weather and has strong penetration ability, data when floods occur can be obtained and recognition accuracy is improved.

Abstract: A landslide recognition method based on Laplacian pyramid remote sensing image fusion includes: performing original remote sensing image reconstruction based on extracted local features and global features of remote sensing images through a Laplacian pyramid fusion module to generate a fused image, constructing a deep learning semantic segmentation model through a semantic segmentation network, labeling the fused image to obtain a dataset of landslide disaster label map, and training the deep learning semantic segmentation model by the dataset, and then storing when a loss curve is fitted and a landslide recognition accuracy of remote sensing image of the deep learning semantics segmentation model meets a requirement by modifying a structure of the semantic segmentation network and adjusting parameters of the deep learning semantics segmentation model.