In this paper, we propose a spectral-spatial feature based classification (SSFC) framework that jointly uses dimension reduction and deep learning techniques for spectral and spatial feature ...extraction, respectively. In this framework, a balanced local discriminant embedding algorithm is proposed for spectral feature extraction from high-dimensional hyperspectral data sets. In the meantime, convolutional neural network is utilized to automatically find spatial-related features at high levels. Then, the fusion feature is extracted by stacking spectral and spatial features together. Finally, the multiple-feature-based classifier is trained for image classification. Experimental results on well-known hyperspectral data sets show that the proposed SSFC method outperforms other commonly used methods for hyperspectral image classification.
It is widely agreed that spatial features can be combined with spectral properties for improving interpretation performances on very-high-resolution (VHR) images in urban areas. However, many ...existing methods for extracting spatial features can only generate low-level features and consider limited scales, leading to unpleasant classification results. In this study, multiscale convolutional neural network (MCNN) algorithm was presented to learn spatial-related deep features for hyperspectral remote imagery classification. Unlike traditional methods for extracting spatial features, the MCNN first transforms the original data sets into a pyramid structure containing spatial information at multiple scales, and then automatically extracts high-level spatial features using multiscale training data sets. Specifically, the MCNN has two merits: (1) high-level spatial features can be effectively learned by using the hierarchical learning structure and (2) multiscale learning scheme can capture contextual information at different scales. To evaluate the effectiveness of the proposed approach, the MCNN was applied to classify the well-known hyperspectral data sets and compared with traditional methods. The experimental results shown a significant increase in classification accuracies especially for urban areas.
Multi-temporal remote sensing imagery has been regarded as an effective tool to monitor cropland. But optical sensors often miss key stages for crop growth because of clouds, which poses challenges ...to many studies. The synergistic of SAR and optical data is expected to lift this problem, especially in areas with persistent cloud cover. However, due to the different characteristics of optical and SAR sensors, it is difficult to build a relationship between the two with most existing methods, let alone construct the long-time correlations to fill optic observation gaps using SAR data. Inspired by deep learning, this study presents a novel strategy to learn the relationship between optical and SAR time series based on the sequence of contextual information. To be specific, we extended the conventional CNN-RNN to build Multi-CNN-Sequence to Sequence (MCNN-Seq) model, and formulate the correlation between the optic and SAR time series sequences. We verified the MCNN-Seq model and found that the accuracy of the predicted optical image was determined by crop types and phenological stages, both in the spatial and temporal domain, respectively. For several crops, such as onion, winter wheat, corn, and sugar beet, our predictions are fitting well with R2 0.9409, 0.9824,0.9157, and 0.9749, respectively. Compared to CNN and RNN, the simulation accuracy achieved by the MCNN-Seq model is much better in terms of R2 and RMSE. In general, results demonstrate that deep learning models have the potential to synergize SAR and optical data and provide replaceable information when the optical data has a long data gap due to the persistent clouds.
•The MCNN-Seq is developed to synergize SAR-Optical data in long time series.•The SAR-NDVI estimator built on SAR-Optical correlation aims to fill NDVI gaps.•The SAR-Optical synergistic is self-adaptive according to crop types.•Experiments on croplands demonstrate its capacity in filling optical values.
•Water adsorption is a feature of dew formation in the revegetation desert ecosystem.•The dew formation relies on different development stages of sand-fixing shrubs.•A mutual enhanced effect exists ...between dew and artificially revegetation system.
Dewfall may be a critical source of moisture in desert environments and may determine sustainability of sand-stabilizing planted vegetation. However, little is known about factors responsible for dew formation, the relative importance of dew as a source of water, and its variability in plantations. During June and October of 2013, the dew amounts and duration were estimated by using the Bowen ratio energy budget technique (BREB), and the dew variability on sand dunes planted with Haloxylon ammodendron 5, 20, and 40 years before were measured by microlysimeter. We quantified dew formation characteristics in a sand-stabilizing H. ammodendron plantation at the edge of a desert oasis, northwestern China. The results indicated that the average daily amount of dew in the H. ammodendron plantations during the observation period based on BREB was 0.13mm, and the dew duration lasted from 1 to 9.5h. Dew occurred on 77% of growing season days, the number of days with dew amounts of >0.03mm comprised 95% of the total dewfall days, and the cumulative amount of dew for those days was 16.1mm. Air temperature, relative humidity, the difference between air temperature and dew point, and wind speed had significant effects on dew formation. The thresholds of the dew formation were RH >50% and wind speed <4.27m/s. As a result of larger canopy area and lower Sky View Factors to 20- and 40-year-old H. ammodendron, the accumulated amount of dew was always significantly greater, and its night-time variability was almost 3 times greater for 5-year-old than for 20- and 40-year-old shrubs. In addition, near-ground dew amounts at the inter-space of three ages of H. ammodendron exhibited higher values than that under the canopy, while dew formation lagged and the maximum cumulative amount of dew was observed 2h later under the canopy of shrubs. The Bowen ratio method estimated actual dew reasonably well. It is concluded that dew may be a frequent and stable water resources in H. ammodendron plantations at the edge of a desert oasis, and there is a mutually reinforcing effect between dewfall and the sand-fixing vegetation system.
Continental shale oil has two types, low-medium maturity and medium-high maturity, and they are different in terms of resource environment, potential, production methods and technologies, and ...industrial evaluation criteria. In addition, continental shale oil is different from the shale oil and tight oil in the United States. Scientific definition of connotations of these resource types is of great significance for promoting the exploration of continental shale oil from “outside source” into “inside source” and making it a strategic replacement resource in the future. The connotations of low-medium maturity and medium-high maturity continental shale oils are made clear in this study. The former refers to the liquid hydrocarbons and multiple organic matter buried in the continental organic-rich shale strata with a burial depth deeper than 300 m and a Ro value less than 1.0%. The latter refers to the liquid hydrocarbons present in organic-rich shale intervals with a burial depth that in the “liquid window” range of the Tissot model and a Ro value greater than 1.0%. The geological characteristics, resource potential and economic evaluation criteria of different types of continental shale oil are systematically summarized. According to evaluation, the recoverable resources of in-situ conversion technology for shale oil with low-medium maturity in China is about (700–900)×108 t, and the economic recoverable resources under medium oil price condition ($ 60–65/bbl) is (150–200)×108 t. Shale oil with low-medium maturity guarantees the occurrence of the continental shale oil revolution. Pilot target areas should be optimized and core technical equipment should be developed according to the key parameters such as the cumulative production scale of well groups, the production scale, the preservation conditions, and the economics of exploitation. The geological resources of medium-high maturity shale oil are about 100×108 t, and the recoverable resources can to be determined after the daily production and cumulative production of a single well reach the economic threshold. Continental shale oil and tight oil are different in lithological combinations, facies distribution, and productivity evaluation criteria. The two can be independently distinguished and coexist according to different resource types. The determination of China's continental shale oil types, resources potentials, and tight oil boundary systems can provide a reference for the upcoming shale oil exploration and development practices and help the development of China's continental shale oil.
In this letter, a novel deep learning framework for hyperspectral image classification using both spectral and spatial features is presented. The framework is a hybrid of principal component ...analysis, deep convolutional neural networks (DCNNs) and logistic regression (LR). The DCNNs for hierarchically extract deep features is introduced into hyperspectral image classification for the first time. The proposed technique consists of two steps. First, feature map generation algorithm is presented to generate the spectral and spatial feature maps. Second, the DCNNs-LR classifier is trained to get useful high-level features and to fine-tune the whole model. Comparative experiments conducted over widely used hyperspectral data indicate that DCNNs-LR classifier built in this proposed deep learning framework provides better classification accuracy than previous hyperspectral classification methods.
In-situ conversion processing (ICP) of shale oil underground at the depth ranging from 300 m to 3 000 m is a physical and chemical process caused by using horizontal drilling and electric heating ...technology, which converts heavy oil, bitumen and various organic matter into light oil and gas in a large scale, which can be called “underground refinery”. ICP has several advantages as in CO2 capture, recoverable resource potential and the quality of hydrocarbon output. Based on the geothermal evolution mechanism of organic materials established by Tissot et al., this study reveals that in the nonmarine organic-rich shale sequence, the amount of liquid hydrocarbon maintaining in the shale is as high as 25% in the liquid hydrocarbon window stage (Ro less than 1.0%), and the unconverted organic materials (low mature-immature organic materials) in the shale interval can reach 40% to 100%. The conditions of organic-rich shale suitable for underground in-situ conversion of shale oil should be satisfied in the following aspects, TOC higher than 6%, Ro ranging between 0.5% and 1%, concentrated thickness of organic-rich shale greater than 15 meters, burial depth less than 3 000 m, covering area bigger than 50 km2, good sealing condition in both up- and down-contacting sequences and water content smaller than 5%, etc. The shale oil resource in China's onshore region is huge. It is estimated with this paper that the technical recoverable resource reaches 70−90 billion tons of oil and 60−65 trillion cubic meters of gas. The ICP of shale oil underground is believed to be a fairway to find big oil in the source kitchen in the near future. And it is also believed to be a milestone to keep China long-term stability of oil and gas sufficient supply by putting ICP of shale oil underground into real practice in the future.
Continental shale oil is a general term for liquid hydrocarbons and many kinds of organic matter in continental organic-rich shale series with vitrinite reflectance of more than 0.5% at buried depth ...of more than 300 m, and is an important type of source-rock oil and gas. Based on the evolution model of oil generation and expulsion in organic-rich shale series controlled by maturity, continental shale oil is divided into two types: medium-high maturity and medium-low maturity. (1) The continental shale series in China develop high-quality source rocks of freshwater and saltwater lacustrine facies, as well as multiple types of reservoirs, including clastic rocks, carbonate rocks, diamictite, tuff and shale, forming a number of “sweet sections” and “sweet areas” of continuous distribution inside or near source rocks, which have large scale resources. (2) Experimental analysis of organic rich shale samples shows that the shale samples with wavy and horizontal beddings have good storage conditions, and the horizontal permeability of shale is tens to hundreds of times of its vertical permeability, which is conducive to the lateral migration and accumulation of shale oil in the source rocks. (3) After evaluation, the geological resources of medium-high maturity shale oil are about 10 billion tons, which can be effectively developed by horizontal drilling and volumetric fracturing, and will be a practical field of oil exploration in recent years. Shale oil with medium and low maturity has huge resource potential, and technological recoverable resources of (70–90) billion tons, making it a strategic alternative resource of oil industry. However, economic development of this type of shale oil needs in-situ conversion technology breakthroughs. Continental shale oil is an inevitable choice in the process of Chinese continental petroleum exploration from “outside source” to “inside source”. Making breakthroughs in the core technologies such as “sweet area” evaluation and optimization, horizontal well volume fracturing and in-situ conversion technology and equipment is the key to realizing scale development of continental shale oil economically.
Overgrazing has caused serious soil degradation, vegetation destruction, and wind erosion in temperate desert areas of northwestern China. To address this, the Livestock Enclosure Scheme was ...implemented in 2004 to allow for self-recovery of the overgrazed desert. The aim of this study was to investigate the effect of livestock enclosure on the restoration process by examining changes in the ecological environment of a temperate desert recovering from overgrazing. Vegetation parameters and soil properties under grazing site, 5-year fenced site, and 9-year fenced site were examined in the degraded desert in the Hexi Corridor of China. The results showed that vegetation coverage, height, plant density, richness, diversity, and litter biomass all improved after 5years of fencing, demonstrating that short-time fencing had a positive effect on vegetation restoration in the temperate desert. Furthermore, 9years of fencing significantly increased soil organic C (SOC), total N (TN), and total P (TP) in the 0–20cm soil profile, while available N (AN) and available P (AP) showed the reverse trend. The values of soil C and N storage after 9years of fencing were almost 5.55Mgha−1 and 0.69Mgha−1, respectively. Enhanced vegetation caused reductions in soil water near the soil surface (0–30cm) where vegetation uptake mainly occurred. These findings could offer insight into the development of effective strategies for protecting and enhancing the resilience of a livestock-disturbed temperate desert.
•Vegetation and soil property response of short-time fencing in temperate desert were studied.•5years of fencing had a positive effect on vegetation restoration.•9years of fencing significantly increased soil nutrients.•The values of soil C and N storage after 9years of fencing were 5.55Mgha−1 and 0.69Mgha−1 in the temperate desert.
Organic-rich marine shales are developed in both the Cambrian Qiongzhusi Formation and the Ordovician Wufeng Formation–Silurian Longmaxi Formation in South China, but are different in the drilling ...results of shale gas exploration. Comparing the differences in shale gas formation conditions between Qiongzhusi and Wufeng-Longmaxi has practical and theoretical significance. This study reveals: (1) in the Sichuan Basin, the Wufeng-Longmaxi Formation has slightly higher TOC than the Qiongzhusi Formation, whereas Qiongzhusi Formation has some local high TOC areas outside of the Sichuan Basin; (2) the Qiongzhusi Formation has much higher thermal evolution degree than the Wufeng-Longmaxi Formation; (3) with undeveloped organic pores, the Qiongzhusi Formation has a 1/3 to 1/2 porosity of the Wufeng-Longmaxi Formation; (4) Qiongzhusi shale has a lower gas content, only 1/2 of that in Wufeng-Longmaxi shale; (5) the Qiongzhusi Formation is mainly composed of siliceous shale and the silica is hot water origin, whereas the Wufeng-Longmaxi Formation consists mainly of calcareous siliceous shale and the silica is biogenic origin; (6) the Wufeng-Longmaxi Formation has overpressure, while the Qiongzhusi Formation is normal in pressure. The reasons for the differences are: (1) different sedimentary environments affect TOC and shale thickness; (2) the Qiongzhusi Formation is over-mature, which caused depletion of hydrocarbon generation, organic carbonization, porosity reduction, and gas content drop; (3) the bad roof and floor conditions of the Qiongzhusi Formation are not good for shale gas preservation; (4) Wufeng-Longmaxi Formation is located in the slope and syncline accompanied with overpressure, and is in favor of preservation and high production of shale gas; (5) the uranium content in the Qiongzhusi Formation is twice that of the Wufeng-Longmaxi Formation, which is the main reason of its higher thermal evolution degree. It is concluded that shale gas enrichment in the marine shale in South China requires favorable geological conditions: organic-rich intervals, moderate thermal evolution, rich organic pores, high gas content, good roof and floor preservation conditions, and moderate burial depth. The Wufeng-Longmaxi Formation has better shale gas enrichment conditions and higher resource potential, whereas the Qiongzhusi Formation has poorer shale gas accumulation conditions and limited favorable areas.
PDF
