Marine oil spills can cause severe damage to the marine environment and biological resources. Using satellite remote sensing technology is one of the best ways to monitor the sea surface in near real-time to obtain oil spill information. The existing methods in the literature either use deep convolutional neural networks in synthetic aperture radar (SAR) images to directly identify oil spills or use traditional methods based on artificial features sequentially to distinguish oil spills from sea surface. However, both approaches currently only use image information and ignore some valuable auxiliary information, such as marine weather conditions, distances from oil spill candidates to oil spill sources, etc. In this study, we proposed a novel method to help detect marine oil spills by constructing a multi-source knowledge graph, which was the first one specifically designed for oil spill detection in the remote sensing field. Our method can rationally organize and utilize various oil spill-related information obtained from multiple data sources, such as remote sensing images, vectors, texts, and atmosphere-ocean model data, which can be stored in a graph database for user-friendly query and management. In order to identify oil spills more effectively, we also proposed 13 new dark spot features and then used a feature selection technique to create a feature subset that was favorable to oil spill detection. Furthermore, we proposed a knowledge graph-based oil spill reasoning method that combines rule inference and graph neural network technology, which pre-inferred and eliminated most non-oil spills using statistical rules to alleviate the problem of imbalanced data categories (oil slick and non-oil slick). Entity recognition is ultimately performed on the remaining oil spill candidates using a graph neural network algorithm. To verify the effectiveness of our knowledge graph approach, we collected 35 large SAR images to construct a new dataset, for which the training set contained 110 oil slicks and 66264 non-oil slicks from 18 SAR images, the validation set contained 35 oil slicks and 69005 non-oil slicks from 10 SAR images, and the testing set contained 36 oil slicks and 36281 non-oil slicks from the remaining 7 SAR images. The results showed that some traditional oil spill detection methods and deep learning models failed when the dataset suffered a severe imbalance, while our proposed method identified oil spills with a sensitivity of 0.8428, specificity of 0.9985, and precision of 0.2781 under those same conditions. The knowledge graph method we proposed using multi-source data can not only help solve the problem of information island in oil spill detection, but serve as a guide for construction of remote sensing knowledge graphs in many other applications as well. The dataset gathered has been made freely available online (https://pan.baidu.com/s/1DDaqIljhjSMEUHyaATDIYA?pwd=qmt6).
Leaf chlorophyll content (LCC) is a key indicator in representing the photosynthetic capacity of Populus deltoides (Populus deltoides Marshall). Unmanned aerial vehicle (UAV) hyperspectral imagery provides an effective approach for LCC estimation, but the issue of band redundancy significantly impacts model accuracy and computational efficiency. Commonly used single feature selection algorithms not only fail to balance computational efficiency with optimal set search but also struggle to combine different regression algorithms under dynamic set conditions. This study proposes an ensemble feature selection framework to enhance LCC estimation accuracy using UAV hyperspectral data. Firstly, the embedded algorithm was improved by introducing the SHapley Additive exPlanations (SHAP) algorithm into the ranking system. A dynamic ranking strategy was then employed to remove bands in steps of 10, with LCC models developed at each step to identify the initial band subset based on estimation accuracy. Finally, the wrapper algorithm was applied using the initial band subset to search for the optimal band subset and develop the corresponding model. Three regression algorithms including gradient boosting regression trees (GBRT), support vector regression (SVR), and gaussian process regression (GPR) were combined with this framework for LCC estimation. The results indicated that the GBRT-Optimal model developed using 28 bands achieved the best performance with R2 of 0.848, RMSE of 1.454 μg/cm2 and MAE of 1.121 μg/cm2. Compared with a model performance that used all bands as inputs, this optimal model reduced the RMSE value by 24.37%. In addition to estimating biophysical and biochemical parameters, this method is also applicable to other hyperspectral imaging tasks.
Abstract Restoration of mangrove forests has garnered increasing global prominence as a nature‐based solution for carbon (C) sequestration. However, it was unclear whether the radiation forcing induced by methane (CH 4 ) emissions and albedo changes during mangrove restoration processes can offset the cooling effect resulting from the net carbon dioxide (CO 2 ) uptake. In this study, we measured the CO 2 , CH 4 , and albedo during 2020–2022 using an open‐path eddy covariance system in an 8‐year restored mangrove ecosystem afforested in Zhejiang Province, China. Their integrated global warming potentials (GWPs) were calculated to assess the climatic impact of mangrove restoration. The results showed that the restored mangroves functioned as a CO 2 sink and a CH 4 source, with annual values of −656.75 to −465.41 and 5.54 to 9.07 g C m −2 yr −1 , respectively. The albedo varied slightly with a range of 0.11–0.13. The integrated GWPs of CO 2 , CH 4 , and albedo were −1,354.00 and −1,875.70 g CO 2 ‐eq m −2 yr −1 over the 20‐ and 100‐year time horizons, respectively. The negative values indicated that the mangrove restoration had a net cooling effect, mainly due to CO 2 uptake. The warming effects caused by CH 4 emissions and albedo changes had the potential to partially offset CO 2 uptake by 12.55%–36.51% and 0.08%–0.42%, respectively. Random forest analysis showed that photosynthetically active radiation (PAR) was the dominant driver on integrated GWPs with feature importance values of 0.34. Our results revealed that the cooling benefit of 8‐year restored mangroves remained significant, even when it was partially offset by CH 4 emissions and albedo changes.
Abstract Blue carbon in mangrove ecosystems contributes significantly to the global carbon cycle. However, large uncertainties maintain in the soil organic carbon (SOC) storage throughout the tide-induced salinity and alkalinity transect in the mangrove restoration region in Southern China. Total 125 soil samples were obtained to detect the SOC content and physicochemical properties. The mean SOC content of each layer ranged from 6.82 to 7.86 g kg −1 , while the SOC density ranged from 2.99 to 11.41 kg m −2 , increasing with soil depths. From different land covers in the study region, the SOC content varied from 4.63 to 9.71 g kg −1 , increasing across the salinity and alkalinity transect, while the SOC density fluctuated from 3.01 kg m −2 in mudflats to 10.05 kg m −2 in mangrove forests. SOC concentration was favorably linked with total nitrogen (r = 0.95), and total phosphorus (r = 0.74), and negatively correlated with Cl − (r = − 0.95), electrical conductivity (r = − 0.24), and total dissolved solids (r = − 0.08). There were significant logarithmic relationships between SOC content and the concentrations of clay (r = 0.76), fine silt (r = 0.81), medium silt (r = − 0.82), and coarse silt (r = − 0.78). The spatial patterns of SOC concentration were notably affected by soil texture, physicochemical properties, and land-cover type, providing essential reference for future investigations of blue carbon budget in restored mangrove forests.
Synthetic Aperture Radar (SAR) is the primary equipment used to detect oil slicks on the ocean’s surface. On SAR images, oil spill regions, as well as other places impacted by atmospheric and oceanic phenomena such as rain cells, upwellings, and internal waves, appear as dark spots. Dark spot detection is typically the initial stage in the identification of oil spills. Because the identified dark spots are oil slick candidates, the quality of dark spot segmentation will eventually impact the accuracy of oil slick identification. Although certain sophisticated deep learning approaches employing pixels as primary processing units work well in remote sensing image semantic segmentation, finding some dark patches with weak boundaries and small regions from noisy SAR images remains a significant difficulty. In light of the foregoing, this paper proposes a dark spot detection method based on superpixels and deeper graph convolutional networks (SGDCNs), with superpixels serving as processing units. The contours of dark spots can be better detected after superpixel segmentation, and the noise in the SAR image can also be smoothed. Furthermore, features derived from superpixel regions are more robust than those derived from fixed pixel neighborhoods. Using the support vector machine recursive feature elimination (SVM-RFE) feature selection algorithm, we obtain an excellent subset of superpixel features for segmentation to reduce the learning task difficulty. After that, the SAR images are transformed into graphs with superpixels as nodes, which are fed into the deeper graph convolutional neural network for node classification. SGDCN leverages a differentiable aggregation function to aggregate the node and neighbor features to form more advanced features. To validate our method, we manually annotated six typical large-scale SAR images covering the Baltic Sea and constructed a dark spot detection dataset. The experimental results demonstrate that our proposed SGDCN is robust and effective compared with several competitive baselines. This dataset has been made publicly available along with this paper.
Abstract Forest disturbances can result in very different canopies that carry elevated albedo, thus causing substantial cooling effects on the climate. Unfortunately, the resulting dynamic global warming potential from altered albedo (GWP Δα ) is poorly understood. We examined and modeled the changes in albedo over time after disturbances (i.e., forest age) by forest type, disturbance type and geographic location using direct measurements from 107 sites in temperate and boreal regions. Albedo in undisturbed forests was used as the reference to calculate albedo changes (Δα) and GWP Δα after a disturbance. We found that age is a significant factor for predicting albedo amid the obvious regulations from forest type and geographic locations. We found the strongest cooling GWP Δα in the first 10 years after a disturbance, but it decreased rapidly with time. The changes in GWP Δα were very different from the chronosequence of net ecosystem production (NEP). In the first decade after disturbances, GWP Δα was negative (i.e., cooling) and surprisingly larger in magnitude, with an average of −0.609 kg CO 2 m −2 yr −1 , compared to NEP of −0.166 kg CO 2 m −2 yr −1 . Albedo continued to decrease and approached pre‐disturbance levels until around 50 years, resulting in a nearly zero GWP Δα . This research illustrates that many forests in temperate and boreal regions can be considered significant cooling agents by taking into account the high albedo of young forests following disturbances.
'/%3e%3cuse xlink:href='%23a' transform='rotate(23.036 .093 25.536)'/%3e%3cuse xlink:href='%23a' transform='rotate(45.87 1.273 16.18)'/%3e%3cuse xlink:href='%23a' transform='rotate(69.945 .996 12.078)'/%3e%3cuse xlink:href='%23a' transform='rotate(20.66 -19.689 31.932)'/%3e%3c/svg%3e)
