Land use and land cover (LULC) changes resulting from rapid urbanization are the foremost causes of increases in land surface temperature (LST) in urban areas. Exploring the impact of LULC changes on the spatiotemporal patterns of LST under future climate change scenarios is critical for sustainable urban development. This study aimed to project the LST of Nanjing for 2025 and 2030 under different climate change scenarios using simulated LULC and land coverage indicators. Thermal infrared data from Landsat images were used to derive spatiotemporal patterns of LST in Nanjing from 1990 to 2020. The patch-generating land use simulation (PLUS) model was applied to simulate the LULC of Nanjing for 2025 and 2030 using historical LULC data and spatial driving factors. We simulated the corresponding land coverage indicators using simulated LULC data. We then generated LSTs for 2025 and 2030 under different climate change scenarios by applying regression relationships between LST and land coverage indicators. The results show that the LST of Nanjing has been increasing since 1990, with the mean LST increased from 23.44 °C in 1990 to 25.40 °C in 2020, and the mean LST estimated to reach 26.73 °C in 2030 (SSP585 scenario, integrated scenario of SSP5 and RCP5.8). There were significant differences in the LST under different climate scenarios, with increases in LST gradually decreasing under the SSP126 scenario (integrated scenario of SSP1 and RCP2.6). LST growth was similar to the historical trend under the SSP245 scenario (integrated scenario of SSP2 and RCP4.5), and an extreme increase in LST was observed under the SSP585 scenario. Our results suggest that the increase in impervious surface area is the main reason for the LST increase and urban heat island (UHI) effect. Overall, we proposed a method to project future LST considering land use change effects and provide reasonable LST scenarios for Nanjing, which may be useful for mitigating the UHI effect.

Forest biomass is closely related to carbon sequestration capacity and can reflect the level of forest management. This study utilizes four machine learning algorithms, namely Multivariate Stepwise Regression (MSR), K-Nearest Neighbors (k-NN), Artificial Neural Network (ANN), and Random Forest (RF), to estimate forest aboveground biomass (AGB) in Chenzhou City, Hunan Province, China. In addition, a canopy height model, constructed from a digital surface model (DSM) derived from Sentinel-1 Interferometric Synthetic Aperture Radar (InSAR) and an ICESat-2-corrected SRTM DEM, is incorporated to quantify its impact on the accuracy of AGB estimation. The results indicate the following: (1) The incorporation of multi-source remote sensing data significantly improves the accuracy of AGB estimation, among which the RF model performs the best (R² = 0.69, RMSE = 24.26 t·ha⁻¹) compared with the single-source model. (2) The canopy height model (CHM) obtained from InSAR-LiDAR effectively alleviates the signal saturation effect of optical and SAR data in high-biomass areas (>200 t·ha⁻¹). When FCH is added to the RF model combined with multi-source remote sensing data, the R² of the AGB estimation model is improved to 0.74. (3) In 2018, AGB in Chenzhou City shows clear spatial heterogeneity, with a mean of 51.87 t·ha⁻¹. Biomass increases from the western hilly part (32.15–68.43 t·ha⁻¹) to the eastern mountainous area (89.72–256.41 t·ha⁻¹), peaking in Dongjiang Lake National Forest Park (256.41 t·ha⁻¹). This study proposes a comprehensive feature integration framework that combines red-edge spectral indices for capturing vegetation physiological status, SAR-derived texture metrics for assessing canopy structural heterogeneity, and canopy height metrics to characterize forest three-dimensional structure. This integrated approach enables the robust and accurate monitoring of carbon storage in subtropical forests.

Evaluating the long-term urban forest ecological environmental quality (EEQ) and analyzing the drivers of its spatiotemporal changes can provide a scientific basis for making long-term urban forest planning decisions. Taking into account the characteristics of urban forest parks with low area proportions of construction land and bare land, high vegetation coverage, and serious forest disturbances, we constructed a modified urban forest park EEQ evaluation index based on a remote sensing ecological index named MRSEI, which is composed of the Landsat enhanced vegetation index (EVI), wetness, land surface temperature (LST), and forest disturbance index (FDI). We selected the Nanjing Zijin Mountain National Forest Park as the study area, used landsat time series stack (LTSS) remote sensing images from 1990 to 2020 as the main data source, and adopted the suggested modified MRSEI, the Theil-Sen median method, and the Hurst index to evaluate the EEQ to analyze its spatiotemporal variations and its driving factors in the study area. The main research results were as follows: (1) the EEQ of Zijin Mountain showed an up-and-down, overall slowly increasing trend from 1990 to 2020, while the spatial auto-correlation coefficient showed an overall decreasing trend; (2) the area percentage of the EEQ-persistent region accounted for 78.69%, and the anti-sustainable region accounted for 21.31%; (3) the spatial centers of the EEQ in the study area were mainly concentrated on the middle and upper part of the southern slope of Zijin Mountain, moving southward from 1990 to 2020; (4) the analysis of drivers showed that climate factors, forest landscape structure, forest disturbances, and forest growth conditions were the main driving factors affecting the EEQ in the study area. These results provide a research framework for the analysis of EEQ changes over a long-term period in the urban forest parks of China.

Efficient image recognition is important in crop and forest management. However, it faces many challenges, such as the large number of plant species and diseases, the variability of plant appearance, and the scarcity of labeled data for training. To address this issue, we modified a SOTA Cross-Domain Few-shot Learning (CDFSL) method based on prototypical networks and attention mechanisms. We employed attention mechanisms to perform feature extraction and prototype generation by focusing on the most relevant parts of the images, then used prototypical networks to learn the prototype of each category and classify new instances. Finally, we demonstrated the effectiveness of the modified CDFSL method on several plant and disease recognition datasets. The results showed that the modified pipeline was able to recognize several cross-domain datasets using generic representations, and achieved up to 96.95% and 94.07% classification accuracy on datasets with the same and different domains, respectively. In addition, we visualized the experimental results, demonstrating the model’s stable transfer capability between datasets and the model’s high visual correlation with plant and disease biological characteristics. Moreover, by extending the classes of different semantics within the training dataset, our model can be generalized to other domains, which implies broad applicability.