In this study, we implement a new frozen‐soil parameterization scheme into the climate system model CAS‐FGOALS‐g3 to investigate the dynamic changes of freezing and thawing fronts and the effects arising from thermal processes and climate. Simulations are conducted using the developed model to validate its performance relative to multi‐source observations. It is shown that the model could reasonably reproduce soil freezing and thawing processes, including dynamic changes in freezing and thawing fronts. The historical simulation shows that the maximum freeze depth increases with an increase of latitude in seasonally frozen ground, and the active layer thickness decreases with an increase of latitude in permafrost regions. The active layer thickness shows increasing trends while the maximum freeze depth shows decreasing trends, which is consistent with change in the 2‐m air temperature. In conclusion, these results have the potential to further deepen our understanding of the freeze‐thaw cycle process and the historical response of permafrost to climate change.

This study investigates the responses of the urban atmospheric thermal environment to two distinct heat waves in Hefei, China, and explores potential changes associated with future urban expansion. During the Event 1, characterized by clear and dry conditions, the western Pacific subtropical high limits water vapor influx, resulting in a significant cooling effect in rural area due to higher surface latent heat flux. The urban heat island (UHI) intensity, calculated using surface temperature and 2‐m temperature, reaches 5.2°C and 1.7°C during the Event 1, respectively. Although Event 2, characterized by cloudy and humid conditions, exhibits weaker UHI and urban dry island effects, it remains highly unfavorable for human comfort. During distinct heat waves, the vertical extent of the warming effect induced by future urban expansion varies, which can be attributed to environmental factors, such as atmospheric stability and near‐surface wind speed.

The rise of knowledge graphs has been instrumental in advancing artificial intelligence (AI) research. Extracting entity and relation triples from unstructured text is crucial for the construction of knowledge graphs. However, Chinese text has a complex grammatical structure, which may lead to the problem of overlapping entities. Previous pipeline models have struggled to address such overlap problems effectively, while joint models require entity annotations for each predefined relation in the set, which results in redundant relations. In addition, the traditional models often lead to task imbalance by overlooking the differences between tasks. To tackle these challenges, this research proposes a global pointer network based on relation prediction and loss function improvement (GPRL) for joint extraction of entities and relations. Experimental evaluations on the publicly available Chinese datasets DuIE2.0 and CMeIE demonstrate that the GPRL model achieves a 1.2–26.1% improvement in F1 score compared with baseline models. Further, experiments of overlapping classification conducted on CMeIE have also verified the effectiveness of overlapping triad extraction and ablation experiments. The model is helpful in identifying entities and relations accurately and can reduce redundancy by leveraging relation filtering and the global pointer network. In addition, the incorporation of a multi-task learning framework balances the loss functions of multiple tasks and enhances task interactions.

Currently, land drought events have become a frequent and serious global disaster. How to address these droughts has become a major issue for researchers. Traditional response strategies for land drought events have been determined by experts based on the severity levels of the events. However, these methods do not account for temporal variations or the specific risks of different areas. As a result, they overlooked the importance of spatio-temporal multi-scale strategies. This research proposes a multi-scale response strategy recommendation model for land drought events. The model integrates characteristics of drought-causing factors, disaster-prone environments, and hazard-bearing bodies using case-based reasoning (CBR). Additionally, the analytic hierarchy process (AHP) and entropy weighting methods (EWMs) are introduced to assign weights to the feature attributes. A case retrieval algorithm is developed based on the similarity of these attributes and the structural similarities of drought cases. The research further classifies emergency strategies into long-term and short-term approaches. Each approach has a corresponding correction algorithm. For short-term strategies, a correction algorithm based on differential evolutions is applied. For long-term strategies, a correction algorithm based on drought risk assessment is developed. The algorithm considers factors such as drought risk, vulnerability, and exposure. It facilitates multi-scale decision-making for drought events. The candidate case obtained using the correction algorithm shows an overall attribute similarity of 94.7% with the real case. The emergency response levels match between the two cases. However, the funding required in the candidate case is CNY 327 million less than the actual expenditure.