Oil tank inventory is significant for the economy and the military, as it can be used to estimate oil reserves. Traditional oil tank detection methods mainly focus on the geometrical characteristics and spectral features of remotely sensed images based on feature engineering. The methods have a limited application capability when the distribution pattern of ground objects in the image changes and the imaging condition varies largely. Therefore, we propose an end-to-end deep convolution network Res2-Unet+, to detect oil tanks in a large-scale area. The Res2-Unet+ method replaces the typical convolution block in the encoder of the original Unet method using hierarchical residual learning branches. A hierarchical branch is used to decompose the feature map into a few sub-channel features. To evaluate the generalization and transferability of the proposed model, we use high spatial resolution images from three different sensors in different areas to train the oil tank detection model. Images from yet another sensor in another area are used to evaluate the trained model. Three more widely used methods, Unet, Segnet, and PSPNet, are trained and evaluated for the same dataset. The experiments prove the effectiveness, strong generalization, and transferability of the proposed Res2-Unet+ method.

With the seriously polluted environment due to social development, the sustainability of net primary production (NPP), which is used to feed most lives on the earth, has become one of the biggest concerns that we have to consider for the sake of food shortage. There have been many researches analyzing one or two potential impact factors of NPP based on field observation data, which brings about many uncertainties for further calculation. Moreover, the frequently used process-based models heavily depend on the understandings of researchers about the NPP process. The premises of such models hinder the impact factor analysis from being objective and confident. To overcome such shortages, we collected 27 potential impact factors of global NPP in terms of eight land cover types. The feature variables include atmosphere, biosphere, anthroposphere and lithosphere parameters, which can be obtained from public available remote sensed products. The experiment shows that latitude, irradiance ultraviolet and normalized difference vegetation index are dominant factors impacting global NPP. Anthropogenic activities, precipitation and surface emissivity are influencing NPP calculation largely. However, some commonly used biosphere parameters in process-based models are actually not playing that important roles in NPP estimation. This work provides a new insight in analyzing NPP impact factors, being more objective and comprehensive compared with frequently used process-based models.

Net primary production (NPP) is one of the fundamental sources of energy for human lives. Due to the tremendous speed of population growth and industrialization, the anthropogenic impact on natural resources is becoming more and more apparent. However, most methods in analyzing potential impact factors of NPP among biosphere parameters are based on process-based models using experimental data mostly collected in the field, which are quite subjective and difficult to share. The process-based models are limited by various premises due to various understandings in constructing the model. To objectively understand the impact factors on calculating global NPP, we build a data-driven model to calculate NPP and analyze each feature variable importance. The features cover biosphere parameters, solar radiation parameters, anthropogenic activities, and atmospheric gas concentrations. The study shows our data-driven model is capable to calculate NPP comparatively well than the process-based model. It further suggests that anthropogenic activities, land cover type, and normalized vegetation index should be considered more in the future NPP-relevant researches, compared with the commonly used biosphere features, such as aerosol index, soil moisture, and precipitation.

Methane (CH₄) is a significant greenhouse gas in exacerbating climate change. Approximately 25 % of CH₄ is emitted from storage tanks. It is crucial to spatially explore the CH₄ emission patterns from storage tanks for efficient strategy proposals to mitigate climate change. However, due to the lack of publicly accessible storage tank locations and distributions, it is difficult to ascertain the CH₄ emission spatial pattern over a large-scale area. To address this problem, we generated a storage tank dataset (STD) by implementing a deep learning model with manual refinement based on high spatial resolution images (1–2m) from the GaoFen-1, GaoFen-2, GaoFen-6, and Ziyuan-3 satellites over cities in China with officially reported numerous storage tanks in 2021. STD is the first storage tank dataset over 92 typical cities in China. The dataset can be accessed at https://zenodo.org/records/10514151 (Chen et al., 2024). It provides a detailed georeferenced inventory of 14,461 storage tanks, wherein each storage tank is validated and assigned the construction year (2000–2021) by visual interpretation referring to the collected high spatial resolution images, historical high spatial resolution images of Google Earth, and field survey. The inventory comprises storage tanks having various distribution patterns in different cities. Spatial consistency analysis with CH₄ emission product shows good agreement with storage tank distributions. The intensive construction of storage tanks significantly induces CH₄ emissions from 2005 to 2020, underscoring the need for more robust measures to curb CH₄ release and aid in climate change mitigation efforts. Our proposed dataset STD will foster the accurate estimation of CH₄ released from storage tanks for CH₄ control and reduction and ensure more efficient treatment strategies are proposed to better understand the impact of storage tanks on the environment, ecology, and human settlements.

Methane (CH₄) is a significant greenhouse gas in exacerbating climate change. Approximately 25 % of CH₄ is emitted from storage tanks. It is crucial to spatially explore the CH₄ emission patterns from storage tanks for efficient strategy proposals to mitigate climate change. However, due to the lack of publicly accessible storage tank locations and distributions, it is difficult to ascertain the CH₄ emission spatial pattern over a large-scale area. To address this problem, we generated a storage tank dataset (STD) by implementing a deep learning model with manual refinement based on high spatial resolution images (1–2m) from the GaoFen-1, GaoFen-2, GaoFen-6, and Ziyuan-3 satellites over cities in China with officially reported numerous storage tanks in 2021. STD is the first storage tank dataset over 92 typical cities in China. The dataset can be accessed at https://zenodo.org/records/10514151 (Chen et al., 2024). It provides a detailed georeferenced inventory of 14,461 storage tanks, wherein each storage tank is validated and assigned the construction year (2000–2021) by visual interpretation referring to the collected high spatial resolution images, historical high spatial resolution images of Google Earth, and field survey. The inventory comprises storage tanks having various distribution patterns in different cities. Spatial consistency analysis with CH₄ emission product shows good agreement with storage tank distributions. The intensive construction of storage tanks significantly induces CH₄ emissions from 2005 to 2020, underscoring the need for more robust measures to curb CH₄ release and aid in climate change mitigation efforts. Our proposed dataset STD will foster the accurate estimation of CH₄ released from storage tanks for CH₄ control and reduction and ensure more efficient treatment strategies are proposed to better understand the impact of storage tanks on the environment, ecology, and human settlements.