Accurately predicting winter wheat yield before harvest could greatly benefit decision-makers when making management decisions. In this study, we utilized weather forecast (WF) data combined with Sentinel-2 data to establish the deep-learning network and achieved an in-season county-scale wheat yield prediction in China’s main wheat-producing areas. We tested a combination of short-term WF data from the China Meteorological Administration to predict in-season yield at different forecast lengths. The results showed that explicitly incorporating WF data can improve the accuracy in crop yield predictions [Root Mean Square Error (RMSE) = 0.517 t/ha] compared to using only remote sensing data (RMSE = 0.624 t/ha). After comparing a series of WF data with different time series lengths, we found that adding 25 days of WF data can achieve the highest yield prediction accuracy. Specifically, the highest accuracy (RMSE = 0.496 t/ha) is achieved when predictions are made on Day of The Year (DOY) 215 (40 days before harvest). Our study established a deep-learning model which can be used for early yield prediction at the county level, and we have proved that weather forecast data can also be applied in data-driven deep-learning yield prediction tasks.

The maximum light use efficiency (LUE) (ε₀) is a key essential parameter of the LUE model, and its accurate estimation is crucial for quantifying gross primary production (GPP) and better understanding the global carbon budget. Currently, a comprehensive understanding of the potential of seasonal variations of ε₀ in GPP estimation across different plant functional types (PFTs) is still lacking. In this study, we used a phenology-based strategy for the estimation of ε₀ to find the optimal photosynthetic responses of the parameter in different phenological stages. The start and end of growing season (SOS and EOS) from time series vegetation indices and the camera-derived greenness index were extracted across seven PFT flux sites using the methods of the hybrid generalized additive model (HGAM) and double logistic function (DLF). Optimal extractions of SOS and EOS were evaluated, and the ε₀ was estimated from flux site observations during the optimal phenological stages with the light response equation. Coupled with other obligatory parameters of the LUE model, phenology-based GPP (GPP_phe-based) was estimated over 21 site-years and compared with vegetation photosynthesis model (VPM)-based GPP (GPP_VPM) and eddy covariance-measured GPP (GPP_EC). Generally, GPP_phe-based basically tracked both the seasonal dynamics and inter-annual variation of GPP_EC well, especially at forest, cropland, and wetland flux sites. The R² between GPP_phe-based and GPP_EC was stable between 0.85 and 0.95 in forest ecosystems, between 0.75 and 0.85 in cropland ecosystems, and around 0.9 in wetland ecosystems. Furthermore, we found that GPP_phe-based was significantly improved compared to GPP_VPM in cropland, grassland, and wetland ecosystems, implying that phenology-based ε₀ is more appropriate in the GPP estimation of herbaceous plants. In addition, we found that GPP_phe-based was significantly improved over GPP_VPM in cropland, grassland, and wetland ecosystems, and the R² between GPP_phe-based and GPP_EC was improved by up to 0.11 in cropland ecosystems and 0.05 in wetland ecosystems compared to GPP_VPM, and RMSE was reduced by up to 5.90 and 2.11 g C m⁻² 8 day⁻¹, respectively, implying that phenology-based ε₀ in herbaceous plants is more appropriate for GPP estimation. This work highlights the potential of phenology-based ε₀ in understanding the seasonal variation of vegetation photosynthesis and production.

Urban forest tree species composition and spatial distribution are essential for refined greenspace management, ecosystem-service assessment, and forest-health monitoring. High-resolution imagery captures crown texture and spatial boundaries, multi-temporal NDVI reflects phenological differences, LiDAR provides canopy height and structural information, and hyperspectral imagery offers detailed spectral discrimination. However, differences in spatial resolution and information representation limit the effectiveness of simple feature stacking. Motivated by the challenges associated with heterogeneous stand structure, similarity among broadleaved species, and insufficient multi-source fusion, this study developed HybridFusion for fine-grained tree species classification in the northern section of Beijing Olympic Forest Park. The model uses 0.5 m high-resolution imagery as its spatial basis and integrates NDVI phenological features from March, April, and August, LiDAR-derived canopy structural features, and hyperspectral–semantic priors through branch encoding, dynamic gated fusion, and reliability-aware hyperspectral modulation. The full-modality model achieved optimal performance (OA: 90.44%; Tree OA: 85.06%; mIoU: 77.88%). Poplar achieved an F1 score of 95.31%, while willow, coniferous trees, and other tree species reached 86.27%, 84.90%, and 85.35%, respectively. Ablation studies confirmed the complementary value of the multi-source data: integrating NDVI and LiDAR increased Tree OA from 67.29% to 81.90%, and the addition of hyperspectral–semantic priors further raised it to 85.06%. These results demonstrate the effectiveness of jointly exploiting phenological, structural, and spectral–semantic information for fine-scale urban forest inventory and management.