Liheng Zhong

Carbon neutrality has become a global priority, and high spatio-temporal resolution data on the column-average dry-air mole fraction carbon dioxide (XCO₂) is essential for tracking progress and guiding policy adjustments. However, satellite-derived XCO₂ exhibits significant temporal and spatial gaps due to influences such as orbital dynamics and cloud cover. Additionally, the low spatial resolution of CarbonTracker (CT) is insufficient to meet the current demands for fine-scale monitoring. Current XCO₂ assessment methods often rely solely on single-pixel data, overlooking the spatio-temporal correlations. In this article, we introduce a deep learning-based spatio-temporal model (DSTM) that extracts features from multiple data sources related to atmospheric transport, carbon emissions, and carbon sinks, enabling fine-scale XCO₂ assessments. Additionally, XCO₂ data from the Orbiting Carbon Observatory-2 (OCO-2) and CT were fused at a 0.1° spatial resolution to generate training labels with broader coverage and more samples, serving as fitting labels. Our approach produced daily, full-coverage 0.1° resolution XCO₂ maps for China from 2015 to 2020, and analyzed changes in XCO₂ growth trends over this period. Numerical results show that our model outperforms traditional deep learning methods. Model validation using data from four ground-based observation sites of the Total Carbon Column Observing Network (TCCON) achieved an average R² of 0.86 and an RMSE of 2.67 ppm. The extraction and fusion of spatio-temporal features from multiple data sources provide a novel approach for reconstructing missing XCO₂ data.

The speckle noise inherent in synthetic aperture radar (SAR) imaging has long posed a challenge for SAR data processing, significantly affecting image interpretation and recognition. Recently, deep learning-based SAR speckle removal algorithms have shown promising results. However, most existing algorithms rely on convolutional neural networks (CNN), which may struggle to effectively capture global image information and lead to texture loss. Besides, due to the different characteristics of optical images and synthetic aperture radar (SAR) images, the results of training with simulated SAR data may bring instability to the real-world SAR data denoising. To address these limitations, we propose an innovative approach that integrates swin transformer blocks into the prediction noise network of the denoising diffusion probabilistic model (DDPM). By harnessing DDPM’s robust generative capabilities and the Swin Transformer’s proficiency in extracting global features, our approach aims to suppress speckle while preserving image details and enhancing authenticity. Additionally, we employ a post-processing strategy known as pixel-shuffle down-sampling (PD) refinement to mitigate the adverse effects of training data and the training process, which rely on spatially uncorrelated noise, thereby improving its adaptability to real-world SAR image despeckling scenarios. We conducted experiments using both simulated SAR image datasets and real SAR image datasets, evaluating our algorithm from subjective and objective perspectives. The visual results demonstrate significant improvements in noise suppression and image detail restoration. The objective results demonstrate that our method obtains state-of-the-art performance, which outperforms the second-best method by an average peak signal-to-noise ratio (PSNR) of 0.93 dB and Structural Similarity Index (SSIM) of 0.03, affirming the effectiveness of our approach.

As one of the most important application for remote sensing monitoring, land cover mapping has witnessed notable advancements in data acquisition, algorithmic diversity, and classification accuracy. Despite the instrumental role data-driven algorithms have played in the development of global land cover products, their inherent limitations as “black box” methods often fall short of meeting end-users’ specific requirements. In this study, built upon the foundation of the earlier land cover monitoring platform [FROM-GLC plus(FGP)], a data and knowledge dual-driven framework (FGP 2.0) was developed as a user-adaptive framework for intelligent remote sensing land cover mapping. By incorporating ontology-based semantic descriptions with advanced data-driven algorithms, FGP 2.0 provides the capacity for both traditional annual mapping and emerging dynamic mapping. Our results illustrate that FGP 2.0 significantly improves the overall accuracy of annual maps by ~5%, and dynamic maps by ~20% compared to FGP. Moreover, an operational dynamic mapping tool has been developed on the Google Earth engine (GEE), enabling the generation of near-real-time land cover maps for any given place. With an extensible and flexible mapping framework, FGP 2.0 demonstrates the potential of customized land cover monitoring results to suit different application scenarios. This innovative approach not only meets the current demand for reliable annual and dynamic land cover maps but also sets a new benchmark for the integration of geoscientific expertise with machine learning techniques in remote sensing monitoring.

Accurately estimating vegetation biomass in urban forested areas is of great interest to researchers as it is a key indicator of the carbon sequestration capacity necessary for cities to achieve carbon neutrality. The emerging vegetation biomass estimation methods that use AI technologies with remote sensing images often suffer from arge estimating errors due to the diversity of vegetation and the complex three-dimensional terrain environment in urban ares. However, the high resolution of Light Detection and Ranging (i.e., LiDAR) data provides an opportunity to accurately describe the complex 3D scenes of urban forests, thereby improving estimation accuracy. Additionally, deep earning foundation models have widely succeeded in the industry, and show great potential promise to estimate vegetation biomass through processing complex and arge amounts of urban LiDAR data efficiently and accurately. In this study, we propose an efficient and accurate method called 3D-CiLBE (3D City Long-term Biomass Estimation) to estimate urban vegetation biomass by utilizing advanced deep earning foundation models. In the 3D-CiLBE method, the Segment Anything Model (i.e., SAM) was used to segment single wood information from a arge amount of complex urban LiDAR data. Then, we modified the Contrastive Language–Image Pre-training (i.e., CLIP) model to identify the species of the wood so that the classic anisotropic growth equation can be used to estimate biomass. Finally, we utilized the Informer model to predict the biomass in the ong term. We evaluate it in eight urban areas across the United States. In the task of identifying urban greening areas, the 3D-CiLBE achieves optimal performance with a mean Intersection over Union (i.e., mIoU) of 0.94. Additionally, for vegetation classification, 3D-CiLBE achieves an optimal recognition accuracy of 92.72%. The estimation of urban vegetation biomass using 3D-CiLBE achieves a Mean Square Error of 0.045 kg/m², reducing the error by up to 8.2% compared to 2D methods. The MSE for biomass prediction by 3D-CiLBE was 0.06kg/m² smaller on average than the inear regression model. Therefore, the experimental results indicate that the 3D-CiLBE method can accurately estimate urban vegetation biomass and has potential for practical application.

Modeling in computer vision has long been dominated by convolutional neural networks (CNNs). Recently, in light of the excellent performance of self-attention mechanism in the language field, transformers tailored for visual data have drawn significant attention and triumphed over CNNs in various vision tasks. These vision transformers heavily rely on large-scale pre-training to achieve competitive accuracy, which not only hinders the freedom of architectural design in downstream tasks like object detection, but also causes learning bias and domain mismatch in the fine-tuning stages. To this end, we aim to get rid of the “pre-train and fine-tune” paradigm of vision transformer and train transformer based object detector from scratch. Some earlier works in the CNNs era have successfully trained CNNs based detectors without pre-training, unfortunately, their findings do not generalize well when the backbone is switched from CNNs to a vision transformer. Instead of proposing a specific vision transformer based detector, in this work, our goal is to reveal the insights of training vision transformer based detectors from scratch. In particular, we expect those insights to help other researchers and practitioners, and inspire more interesting research in other fields, such as remote sensing, visual-linguistic pre-training, etc. One of the key findings is that both architectural changes and more epochs play critical roles in training vision transformer based detectors from scratch. Experiments on the MS COCO dataset demonstrate that vision transformer based detectors trained from scratch can also achieve similar performance to their counterparts with ImageNet pre-training.