Lane-level road maps are crucial for urban traffic management, autonomous driving, and vehicle navigations. Optical remote sensing (ORS) image suffers from trees and buildings occlusion for lane-level road mapping due to the top-down view. While street view images (SVIs) have been used for road detection, however, most of the previous studies focused on extracting road in image space. The reconstruction of lane-level road maps with measurability in geographic space remains challenging. Hence, this paper proposed an operational framework for extracting and reconstructing lane-level road maps from urban open access data (U-OAD). First, a sample strategy was used to collect SVIs based on OSM road central lines. Then, a deep learning (DL)-based method was adopted to identify lanes accurately, and road width was extracted based on design knowledge and OSM information. Finally, the lane-level road map was reconstructed by integrating the lane and its width information. The proposed framework achieves the transformation from image space to geographic space. The case study shows that 82.43% of the roadway is accurately reconstructed in lane-level. The difference between the reconstructed width of the roadway and the reference true value is within the m-level and the RMSE is 0.32 m. The proposed method is cost-effective and accurate-acceptable for acquiring lane-level road datasets in cities.

Determining the amount of nitrogen (N) and phosphorus (P) discharged into lakes and assessing the effectiveness of reduction measures in a basin that involve the interaction of multiple factors are still daunting challenges. In this study, the random forest (RF) model was employed to simulate the impact of controlling measures on the amount of N and P discharged in 2025 under seven specific reduction scenarios. Discharged N and P decreased in the basin by 23.38% and 31.69% from 2011 to 2020, respectively. The N and P nutrient discharge intensities were significantly higher in the western part of the basin (13.31 kg·ha⁻¹·a⁻¹ and 1.34 kg·ha⁻¹·a⁻¹) than those in the eastern region (10.24 kg·ha⁻¹·a⁻¹ and 0.74 kg·ha⁻¹·a⁻¹). Fertilizer runoff (N: 34.72%, 5934.49 t; P: 13.60%, 199.76 t), domestic sewage (N: 29.14%, 4009.27 t; P: 34.84%, 496.59 t), and livestock farming (N: 15.11%, 2657.50 t; P: 40.05%, 616.05 t) were the key sources of N and P. The RF model shows that (R² > 0.994, p < 0.01) the multi-factor reduction effect is the best, and under this discharge reduction effect, the amount of N and P discharged in 2025 are expected to decrease by 13.79% and 19.42%, respectively, compared with those in 2020. In addition, different key sources in sub-basins might lead to regional differences in the discharge reduction effects of various measures. Ultimately, we recommend that the synergistic treatment of point and non-point sources, using treatments with multiple measures, should be implemented in different regions to reduce the amount of N and P discharged in the Chaohu Lake Basin.

Small water bodies serve as fundamental units and key conduits for material cycling and hydrological processes at the watershed scale. However, quantitatively identifying the mechanisms driving their area evolution under multi-factor coupling in different scenario simulations remains challenging. This study focused on the small water bodies in the Chaohu Lake Basin. An Otsu algorithm was applied to establish a basin-scale database of small water bodies, while the GeoDetector model was integrated to reveal the spatiotemporal driving mechanisms of multi-factor coupling behind their evolution. Furthermore, a Long Short-Term Memory (LSTM)–Transformer model was modified to simulate future scenarios of small water body area dynamics. The results indicated that, from 1995 to 2024, the area of small water bodies in the Chaohu Lake Basin exhibited a fluctuating decreasing trend (wet season: 186–269 km²; dry season: 110–253 km²). In terms of spatial distribution, the small water bodies exhibited an unbalanced distribution pattern characterized by wide dispersion alongside regional clustering. Results from the GeoDetector model revealed that land use type (q = 0.711) and evapotranspiration (q = 0.526) were the dominant drivers of variations in small water body areas. LSTM–Transformer simulations (R² = 0.92, p < 0.01) suggested that, under temperature, precipitation, and land use change scenarios, the small water body areas in the Chaohu Lake Basin will exhibit distinct seasonal variation characteristics, with scenario-dependent differences in fluctuation amplitude and peak–trough timing. These results offer theoretical support for the protection of small water bodies and integrated water resource management in the Chaohu Lake Basin.