Coastal cities in China are frequently hit by tropical cyclones (TCs), which result in tremendous loss of life and property. Even though the capability of numerical weather prediction models to forecast and track TCs has considerably improved in recent years, forecasting the intensity of a TC is still very difficult; thus, it is necessary to improve the accuracy of TC intensity prediction. To this end, we established a series of predictors using the Best Track TC dataset to predict the intensity of TCs in the Western North Pacific with an eXtreme Gradient BOOSTing (XGBOOST) model. The climatology and persistence factors, environmental factors, brainstorm features, intensity categories, and TC months are considered inputs for the models while the output is the TC intensity. The performance of the XGBOOST model was tested for very strong TCs such as Hato (2017), Rammasum (2014), Mujiage (2015), and Hagupit (2014). The results obtained show that the combination of inputs chosen were the optimal predictors for TC intensification with lead times of 6, 12, 18, and 24 h. Furthermore, the mean absolute error (MAE) of the XGBOOST model was much smaller than the MAEs of a back propagation neural network (BPNN) used to predict TC intensity. The MAEs of the forecasts with 6, 12, 18, and 24 h lead times for the test samples used were 1.61, 2.44, 3.10, and 3.70 m/s, respectively, for the XGBOOST model. The results indicate that the XGBOOST model developed in this study can be used to improve TC intensity forecast accuracy and can be considered a better alternative to conventional operational forecast models for TC intensity prediction.

Conventional numerical methods have made significant advances in forecasting tropical cyclone (TC) tracks, using remote sensing data with high spatial and temporal resolutions. However, over the past two decades, no significant improvements have been made with regard to the accuracy of TC intensity prediction, which remains challenging, as the internal convection and formation mechanisms of such storms are not fully understood. This study investigated the relationship between remote sensing data and TC intensity to improve the accuracy of TC intensity prediction. An intensity forecast model for the South China Sea was built using the eXtreme Gradient Boosting (XGBoost) model and FengYun-2 (FY-2) satellite data, environmental data, and best track datasets from 2006 to 2017. First, correlation analysis algorithms were used to extract the TC regions in which the satellite data were best correlated, with TC intensity at lead times of 6, 12, 18, and 24 h. Then, satellite, best track, and environmental data were used as source data to develop three different XGBoost models for predicting TC intensity: model A1 (climatology and persistence predictors + environmental predictors), model A2 (A1 + satellite-based predictors extracted as mean values), and model A3 (A1 + satellite-based predictors extracted by our method). Finally, we analyzed the impact of the FY-2 satellite data on the accuracy of TC intensity prediction using the forecast skill parameter. The results revealed that the equivalent blackbody temperature (TBB) of the FY-2 data has a strong correlation with TC intensity at 6, 12, 18, and 24 h lead times. The mean absolute error (MAE) of model A3 was reduced by 0.47%, 1.79%, 1.91%, and 5.04% in 6, 12, 18, and 24 h forecasts, respectively, relative to those of model A2, respectively, and by 2.73%, 7.58%, 7.64%, and 5.04% in 6, 12, 18, and 24 h forecasts, respectively, relative to those of model A1. Furthermore, the accuracy of TC intensity prediction is improved by FY-2 satellite images, and our extraction method was found to significantly improve upon the traditional extraction method.

The violent and persistent wildfires that broke out along the southeast coast of Australia in 2019 caused a large number of pollutant emissions, which seriously affected air quality and the global climate. The existing two methods for estimating combustion emissions based on burned area and fire radiative power mainly use a medium resolution imaging spectrometer (MODIS) on the Aqua and Terra satellites. However, the low temporal resolution of MODIS and insensitivity to small fires lead to deviation in the estimation of fire emissions. In order to solve this problem, the Visible Infrared Imaging Radiometer Suite (VIIRS) with better performance is adopted in this paper, combined with the fire diurnal cycle information obtained by geostationary satellite Himawari-8, to explore the spatio-temporal model of biomass combustion emissions. Using this, a high-spatial- and -temporal-resolution fire emission inventory was generated for southeastern Australia from November 2019 to January 2020, which aims to fully consider the highly dynamic nature of fires and small fires (low FRP) that are much lower than the MODIS burned area or active fire detection limit, with emphasis on dry matter burned (DMB). We found that during the study period, the fire gradually moved from north to south, and the diurnal cycle of the fire in the study area changed greatly. The peak time of the fire gradually delayed as the fire moved south. Our inventory shows that the DMB in southeast Australia during the study period was about 146 Tg, with major burned regions distributed along the Great Dividing Range, with December 2019 being the main burning period. The total DMB we calculated is 0.5–3.1 times that reported by the GFAS (Global Fire Assimilation System) and 1.5 to 4 times lower than that obtained using the traditional “Burned Area Based Method (FINN)”. We believe that the GFAS may underestimate the results by ignoring a large number of small fires, and that the excessive combustion rate used in the FINN may be a source of overestimation. Therefore, we conclude that the combination of high-temporal-resolution and high-spatial-resolution satellites can improve FRE estimation and may also allow further verification of biomass combustion estimates from different inventories, which are far better approaches for fire emission estimation.

Most models designed to simulate pedestrian dynamical behavior are based on the assumption that human decision-making can be described using precise values. This study proposes a new pedestrian model that incorporates fuzzy logic theory into a multi-agent system to address cognitive behavior that introduces uncertainty and imprecision during decision-making. We present a concept of decision preferences to represent the intrinsic control factors of decision-making. To realize the different decision preferences of heterogeneous pedestrians, the Five-Factor (OCEAN) personality model is introduced to model the psychological characteristics of individuals. Then, a fuzzy logic-based approach is adopted for mapping the relationships between the personality traits and the decision preferences. Finally, we have developed an application using our model to simulate pedestrian dynamical behavior in several normal or non-panic scenarios, including a single-exit room, a hallway with obstacles, and a narrowing passage. The effectiveness of the proposed model is validated with a user study. The results show that the proposed model can generate more reasonable and heterogeneous behavior in the simulation and indicate that individual personality has a noticeable effect on pedestrian dynamical behavior.

Deep learning is currently the mainstream approach for building extraction tasks in remote-sensing imagery, capable of automatically learning features of buildings in imagery and yielding satisfactory extraction results. However, due to the diverse sizes, irregular layouts, and complex spatial relationships of buildings, extracted buildings often suffer from incompleteness and boundary issues. Gaofen-7 (GF-7), as a high-resolution stereo mapping satellite, provides well-rectified images from its rear-view imagery, which helps mitigate occlusions in highly varied terrain, thereby offering rich information for building extraction. To improve the integrity of the edges of the building extraction results, this paper proposes a dual-task network (Edge-UNet, EUnet) based on UNet, incorporating an edge extraction branch to emphasize edge information while predicting building targets. We evaluate this method using a self-made GF-7 Building Dataset, the Wuhan University (WHU) Building Dataset, and the Massachusetts Buildings Dataset. Comparative analysis with other mainstream semantic segmentation networks reveals significantly higher F1 scores for the extraction results of our method. Our method exhibits superior completeness and accuracy in building edge extraction compared to unmodified algorithms, demonstrating robust performance.

Generating 3D path models (with textures) from indoor paths is a good way to improve the visualization performance of 3D indoor path analysis. In this paper, a quaternion-based piecewise 3D modeling method is proposed to automatically generate highly recognizable 3D models for indoor path networks. To create such models, indoor paths are classified into four types of basic elements: corridor, stairs, elevator and node, which contain six kinds of edges and seven kinds of nodes. A quaternion-based method is devised to calculate the coordinates of the designed elements, and a piecewise 3D modeling method is implemented to create the entire 3D indoor path models in a 3D GIS scene. The numerical comparison of 3D scene primitives in different visualization modes indicates that the proposed method can generate detailed and irredundant models for indoor path networks. The result of 3D path analysis shows that indoor path models can improve the visualization performance of a 3D indoor path network by displaying paths with different shapes, textures and colors and that the models can maintain a high rendering efficiency (above 50 frames per second) in a 3D GIS scene containing more than 50,000 polygons and triangles.