High-resolution imagery and deep learning models have gained increasing importance in land-use mapping. In recent years, several new deep learning network modeling methods have surfaced. However, there has been a lack of a clear understanding of the performance of these models. In this study, we applied four well-established and robust deep learning models (FCN-8s, SegNet, U-Net, and Swin-UNet) to an open benchmark high-resolution remote sensing dataset to compare their performance in land-use mapping. The results indicate that FCN-8s, SegNet, U-Net, and Swin-UNet achieved overall accuracies of 80.73%, 89.86%, 91.90%, and 96.01%, respectively, on the test set. Furthermore, we assessed the generalization ability of these models using two measures: intersection of union and F1 score, which highlight Swin-UNet’s superior robustness compared to the other three models. In summary, our study provides a systematic analysis of the classification differences among these four deep learning models through experiments. It serves as a valuable reference for selecting models in future research, particularly in scenarios such as land-use mapping, urban functional area recognition, and natural resource management.

Inter-state cyberattacks are increasingly becoming a major hidden threat to national security and global order. However, current prediction models are often constrained by single-source data due to insufficient consideration of complex influencing factors, resulting in limitations in understanding and predicting cyberattacks. To address this issue, we comprehensively consider multiple data sources including cyberattacks, bilateral interactions, armed conflicts, international trade, and national attributes, and propose an interpretable multimodal data fusion framework for predicting cyberattacks among countries. On one hand, we design a dynamic multi-view graph neural network model incorporating temporal interaction attention and multi-view attention, which effectively captures time-varying dynamic features and the importance of node representations from various modalities. Our proposed model exhibits greater performance in comparison to many cutting-edge models, achieving an F1 score of 0.838. On the other hand, our interpretability analysis reveals unique characteristics of national cyberattack behavior. For example, countries with different income levels show varying preferences for data sources, reflecting their different strategic focuses in cyberspace. This unveils the factors and regional differences that affect cyberattack prediction, enhancing the transparency and credibility of the proposed model.

Scrub typhus has emerged as a life-threatening and increasingly prevalent vector-borne disease. While the spatial and temporal distributions of scrub typhus have been studied by the research community, the main driving factors that influence the spatiotemporal patterns of the disease remain under investigation. Using Guangdong Province as a case study, we combined monthly scrub typhus case data from 2010 to 2019 and environmental and vector-related datasets with a Bayesian hierarchical mixed model to elucidate the spatiotemporal characteristics of the disease. This study revealed that the most highly endemic areas of scrub typhus are concentrated in the western and southern parts of Guangdong Province. A distinct bimodal pattern of scrub typhus was observed, with peaks typically occurring from May to July and October to November. The fitted model indicated that forest, cropland, and chigger mites were positively associated with scrub typhus transmission. Furthermore, climate and vectors were identified as key factors shaping the bimodal seasonal patterns of scrub typhus. Despite data-related limitations, including the treatment of population as a time-invariant variable and the binary simplification of vector suitability, sensitivity analyses confirmed the robustness of the model, offering valuable insights for scrub typhus prevention in Guangdong.

The COVID-19 pandemic profoundly altered socio-economic activities, human behaviors, and crime patterns. However, less is known about how the pandemic and associated restrictions affected cyber-enabled and traditional fraud. Here, we conducted a retrospective analysis using police-recorded crime data in the UK to examine the impact of COVID-19 restrictions and changes in human activity on fraud. Results indicate that following the onset of the lockdown, the number of recorded fraud cases increased by 28.5%, contrasting with traditional property crimes, which dropped by 28.1%. However, the lockdown did not have a significant impact on the long-term trend of fraud. With the lifting of restrictions, fraud gradually regressed to levels approaching those before the pandemic. By inspecting the effects of different government responses and changes in population mobility on various types of fraud, we found that more stringent restrictions were associated with larger increases in most types of cyber-enabled fraud, except for those that rely on offline activities, whereas the impact on traditional fraud was mixed and contingent upon specific opportunity structures. These findings overall align with the assumptions of routine activity theory and provide clear support for its applicability in fraud and cybercrime.

Armed conflict is a dynamic social phenomenon, yet existing research often overlooks its evolving nature. We propose a method to simulate the dynamic transformations of armed conflicts. First, we enhanced the Spatial Conflict Dynamic Indicator (SCDi) by integrating conflict intensity and clustering, which allowed for the distinction of various conflict types. Second, we established transformation rules for the SCDi, quantifying five types of transformations: outbreak, stabilization, escalation, de-escalation, and maintaining peace. Using the random forest algorithm with multiple covariates, we simulated these transformations and analyzed the driving factors. The results reveal a global trend of polarization in armed conflicts over the past 20 years, with an increase in clustered/high-intensity (CH) and dispersed/low-intensity (DL) conflicts. Stable regions of ongoing conflict have emerged, notably in areas like Syria, the border of Afghanistan, and Nepal’s border region. New conflicts are more likely to arise near these zones. Various driving forces shape conflict transformations, with neighboring conflict scenarios acting as key catalysts. The capacity of a region to maintain peace largely depends on neighboring conflict dynamics, while local factors are more influential in other types of transformations. This study quantifies the dynamic process of conflict transformations and reveals detailed changes.

The demand for energy plants is foreseen to grow as worldwide energy and climate policies promote the use of bioenergy for climate change mitigation. To avoid competing with food production, it’s critical to assess future changes in marginal land availability for energy plant development. Using a machine learning method, boosted regression tree, this study modeled potential marginal land resources suitable for cassava under current and different climate change scenarios, based on cassava occurrence records and environmental covariates. The findings revealed that, currently, over 80% of the 1357.24 Mha of available marginal land for cassava cultivation is distributed in Africa and South America. Under three climate change scenarios, by 2030, worldwide suitable marginal land resources were predicted to grow by 39.71Mha, 66.21 Mha, and 39.31Mha for the RCP4.5, RCP6.0, and RCP8.5 scenarios, respectively; by 2050, the potential marginal land suitable for cassava will increase by 38.98Mha, 83.02 Mha, and 55.43Mha, respectively; by 2080, the global marginal land resources were estimated to rise by 40.82 Mha, 99.74 Mha, and 21.87 Mha from now, respectively. Our results highlight the impacts of climate change on potential marginal land resources of cassava across worldwide, which provide the basis for assessing bioenergy potential in the future.