Suspended solids concentration (SSC) is an important indicator of the degree of water pollution. However, when using an empirical or semi-empirical model adapted to some of the inland waters to estimate SSC on unmanned aerial vehicle (UAV)-borne hyperspectral images, the accuracy is often not sufficient. Thus, in this study, we attempted to use the particle swarm optimization (PSO) algorithm to find the optimal parameters of the least-squares support vector machine (LSSVM) model for the quantitative inversion of SSC. A reservoir and a polluted riverway were selected as the study areas. The spectral data of the 36-point and 29-point 400–900 nm wavelength range on the UAV-borne images were extracted. Compared with the semi-empirical model, the random forest (RF) algorithm and the competitive adaptive reweighted sampling (CARS) algorithm combined with partial least squares (PLS), the accuracy of the PSO-LSSVM algorithm in predicting the SSC was significantly improved. The training samples had a coefficient of determination ( $R^2$ ) of 0.98, a root mean square error (RMSE) of 0.68 mg/L, and a mean absolute percentage error (MAPE) of 12.66% at the reservoir. For the polluted riverway, PSO-LSSVM also performed well. Finally, the established SSC inversion model was applied to UAV-borne hyperspectral remote sensing (HRS) images. The results confirmed that the distribution of the predicted SSC was consistent with the observed results in the field, which proves that PSO-LSSVM is a feasible approach for the SSC inversion of UAV-borne HRS images.

In urban thermal environment research (UTE), urban shadows formed by buildings and trees contribute to significant variations in thermal conditions, particularly during the mid-day period. This study investigated the multi-scale effects of indicators, including urban shadows, on UTE, focusing specifically on the mid-day hours. It integrated field temperature measurements and drone aerial data from multiple city blocks. Considering both urban shadows and direct solar radiation, a linear mixed-effects model was employed to study the multi-scale effects of urban morphological indicators. Results showed that: (1) UTE is a multi-scale, multi-factor phenomenon, with thermal effects manifesting at specific scales. Under shadow conditions, smaller scales (10–20 m) of landscape heterogeneity and larger scales (300–400 m) of landscape consistency better explained temperature variations mid-day. Conversely, under direct sunlight, temperature was primarily influenced by larger scales (150–300 m). (2) Trees significantly reduced temperature; 100% tree canopy cover within a 10-m radius reduced air temperatures by approximately 2 °C mid-day. However, there is no significant correlation between temperature and green spaces. (3) Building area and height were significantly correlated with temperature. Specifically, an increase in building area beyond 150 m, especially within a 300-m radius, leads to higher temperatures. Conversely, building height within a 10–20 m range exhibits significant cooling effects. These findings provide crucial reference data for micro-scale UTE investigations during mid-day hours and offer new strategies for urban planning and design.

With the acceleration of global climate change and urbanization, the urban heat island effect has significantly impacted the quality of life of urban residents. Although numerous studies have focused on macro-scale factors such as air temperature, surface albedo, and green space coverage, relatively little attention has been paid to micro-scale factors, such as shading provided by building facades and tree canopy coverage. However, these micro-scale factors play a significant role in enhancing pedestrian thermal comfort. This study focuses on a city community in China, aiming to assess the thermal comfort of urban streets during the summer. Utilizing high-resolution 3D geographic data and street view images extracted from drone data, this study comprehensively considers the mechanisms affecting the urban street thermal environment and the human comfort requirements for shading and greening. By proposing quantitative indicators from multiple scales and dimensions, this study thoroughly quantifies the impact of the surrounding environment, greening, shading effects, buildings, and road design on the thermal comfort of summer streets. The results show that increasing tree canopy coverage by 10 m can significantly reduce the surrounding temperature, and a building layout extending 200 m can regulate temperature. The distribution of shadows at different times significantly affects thermal comfort, while the sky view factor negatively correlates with thermal comfort. Environments with a high green view index enhance visual comfort. This study reveals the specific contributions of different environmental characteristics to street thermal comfort and identifies factors that significantly impact thermal comfort. This provides a scientific basis for urban green space planning and thermal comfort improvement, holding substantial practical significance.

Extracting entities from large volumes of chicken epidemic texts is crucial for knowledge sharing, integration, and application. However, named entity recognition (NER) encounters significant challenges in this domain, particularly due to the prevalence of nested entities and domain-specific named entities, coupled with a scarcity of labeled data. To address these challenges, we compiled a corpus from 50 books on chicken diseases, covering 28 different disease types. Utilizing this corpus, we constructed the CDNER dataset and developed a nested NER model, MFGFF-BiLSTM-EGP. This model integrates the multiple fine-grained feature fusion (MFGFF) module with a BiLSTM neural network and employs an efficient global pointer (EGP) to predict the entity location encoding. In the MFGFF module, we designed three encoders: the character encoder, word encoder, and sentence encoder. This design effectively captured fine-grained features and improved the recognition accuracy of nested entities. Experimental results showed that the model performed robustly, with F1 scores of 91.98%, 73.32%, and 82.54% on the CDNER, CMeEE V2, and CLUENER datasets, respectively, outperforming other commonly used NER models. Specifically, on the CDNER dataset, the model achieved an F1 score of 79.68% for nested entity recognition. This research not only advances the development of a knowledge graph and intelligent question-answering system for chicken diseases, but also provides a viable solution for extracting disease information that can be applied to other livestock species.

Extracting entities from large volumes of chicken epidemic texts is crucial for knowledge sharing, integration, and application. However, Named Entity Recognition (NER) encounters significant challenges in this domain, particularly due to the prevalence of nested entities and domain-specific named entities, coupled with a scarcity of labeled data. To address these challenges, we compiled a corpus from 50 books on chicken diseases, covering 28 different disease types. Utilizing this corpus, we developed a nested NER model, MFGFF-BiLSTM-EGP. This model integrates the Multiple Fine-Grained Feature Fusion (MFGFF) module with a BiLSTM neural network and employs an Efficient Global Pointer (EGP) for predicts the entity location encoding. In MFGFF module we designed three encoders, character encoder, word encoder and sentence encoder, this design effectively captures fine-grained features and improves the recognition accuracy of nested entities. Experimental results show that the model performs robustly, with F1 scores of 91.98%, 73.32%, and 82.54% on the CDNER, CMeEE V2, and CLUENER datasets, respectively, outperforming other commonly used NER models. Specifically, on the CDNER dataset, the model achieved an F1 score of 79.68% for nested entity recognition. This research not only advances the development of knowledge graph and intelligent question-answering system for chicken diseases but also provides a viable solution for extracting disease information that can be applied to other livestock species.

Urban heat mitigation requires indicators that capture both static urban form and dynamic radiative processes at the pedestrian scale. This study developed a Dynamic Radiative Indicator System (DRIS) representing three process dimensions within object-based near-ground thermal units: instantaneous solar exposure, neighborhood obstruction, and cumulative obstruction history. The analysis used 64,025 units derived from 14-cm unmanned aerial vehicle (UAV) thermal imagery in a vegetation-rich urban district of Wuhan, China, and integrated land cover, three-dimensional morphology, and simulated shadow dynamics. Synchronized near-ground measurements and an independent 2022 ground dataset were used for supplementary tests. Random forest models showed that adding static morphology to surface features increased held-out R² from 0.677 to 0.807, while further adding DRIS variables increased R² to 0.880 and reduced RMSE to 3.71 °C. Module-specific models showed that cumulative obstruction history contributed explanatory information beyond instantaneous solar exposure and neighborhood obstruction. In the full model, the entity-based indicator of cumulative obstruction history, ranked first, whereas sky view factor (SVF) fell from third to thirteenth after DRIS variables were introduced. Structurally matched ground-shadow and entity-based obstruction models achieved similar predictive performance on UAV-visible surfaces, with only a small practical difference. In the independent 2022 dataset, both obstruction series retained R² values of 0.840–0.842, while entity-based indicators occupied stronger positions in the variable-importance structure. These findings identify cumulative obstruction history as a key explanatory dimension of near-ground temperature variation and provide a computable framework for fine-scale urban heat-mitigation assessment.