Search

In this study, taking the Jurassic Lianggaoshan Formation (J1l) tight sandstones in the eastern Sichuan Basin as an example, the types and well-logging responses of main sedimentological and diagenetic facies in the lacustrine delta-front are investigated based on summarizing the sedimentary characteristics and reservoir properties. Subsequently, further validation and application are conducted in the study area through machine learning. Research results show that the J1l lacustrine delta-front in the eastern Sichuan Basin mainly develops subaqueous distributary channels and mouth bar sand bodies, exhibiting typical densification reservoirs, with porosity and permeability distributed between 0.48% and 11.24% (av. 3.87%) and 0.0003–0.653 × 10−3 μm2 (av. 0.026 × 10−3 μm2), respectively. Strong compaction and strong cementation are the primary factors leading to densification, whereas chlorite coatings and weak dissolution play constructive roles in preserving some primary pores, creating a small amount of dissolution pores, and enhancing permeability. In terms of manifestation, the pore-throat content with a radius greater than 0.006 μm governs the reservoir quality. Furthermore, five types of diagenetic facies are identified in the J1l subaqueous distributary channels and mouth bars: strong compaction facies (Type I), strong cementation facies (Type II), chlorite-coating and intergranular pore facies (Type III), weak dissolution and intragranular pore facies (Type IV), and medium compaction and cementation facies (Type V). Overall, the thick and coarse-grained subaqueous distributary channels can be considered as the preferred exploration targets for tight oil and gas, with type III and type IV diagenetic facies being the most favorable reservoirs, characterized by well-logging responses of high AC and low GR, DEN, and RT. Based on the fine division of sedimentological and diagenetic facies, establishing well-logging interpretation models and then employing machine learning to achieve sweet spot reservoir prediction can provide valuable insights for tight oil and gas exploration in regions lacking core data.

With a large population and rapid urbanization, there are still many challenges to optimize the ecological-agricultural-urban space. Here, taking Yulin City, situated on the Loess Plateau of China as a case in point, we explored the spatial suitability evaluation of ecological-agricultural-urban space. Building upon the Chinese government’s concept of “resource and environmental carrying capacity and territorial development suitability evaluation” (hereinafter referred to as “double evaluation”), this study applies machine learning to the planning of ecological-agricultural-urban space. It explores an intelligent evaluation method for land space patterns using multi-source data. Based on the random forest (RF) algorithm and geographic information system (GIS), resulting in evaluated spatial patterns for ecological-agricultural-urban in the Yulin area. The results showed the constructed random forest models achieved an accuracy of 93% for ecology, 90% for agriculture, and 92% for urban space in Yulin City on the test dataset. By means of suitability analysis, the results indicated that the extremely important ecological space were predominantly located in the southwestern and eastern regions of the study area, while suitable space for agricultural production were primarily scattered throughout the southeast. In contrast, suitable space for urban construction were concentrated mainly in the central part of the study area. The use of machine learning has proven to be effective in addressing multicollinearity among spatial evaluation factors across three different areas. By eliminating human subjectivity in weight assignment during evaluation, it introduces fresh perspectives for land space planning and status assessment. These findings may offer support for the scientific delineation of ecological-agricultural-urban space (three districts and three lines) in China.

With the ever-growing availability of massive geo-data, deep learning has been widely applied to geoscientific questions such as sedimentary provenance analysis. However, randomly selected initial weights (and also biases) and possible loss of population diversity in traditional neural network learning remain problematic. To address this issue, in this study, we proposed a new deep neural network model by incorporating genetic algorithm (GA) and simulated annealing algorithm into the BP neural network, i.e., the GA-SA-BP model. We then applied this new model to rare earth element (REE) geochemical data of the Liuling Group of the East Qinling Orogen to investigate its provenance. Our results showed that among other deep learning algorithms, the new model presents the best performance with good measuring metrics (e.g., over 85% of accuracy, over 0.82 of F1-macro-average, F1-micro-average, and Kappa coefficient, and smallest (

A rhyolitic tuff-mudstone interbedded rock outcrop with a good rhythm is found in the Yangjiaodong area of Lingshan Island, eastern Shandong Province. The volcanic eruption dynamics process was investigated by using the rhythm of volcanic accumulation method. Combined with the geologic setting of the study area, the parameter deposition rate is 6.5 m/Ma, and the mudstone compaction factor is 0.3 because the study area is the Ri-Qing-Wei Basin, which is the volcanic back-arc basin with insufficient source supply. Based on this, the time span of the analyzed sample was calculated to be 2.24 Ma, which represents the duration of this eruption event. We used software called Acycle to finish this research. The result shows cyclicity of 2.296, 5.78, 8.475, and 15.625 kyr and the scales of 2.17, 2.94, 3.12, and 18.87 m repeatability. Considering that magma eruption is affected by changing melt conditions, the crystallinity, water content, temperature, and pressure are comprehensively considered “activation conditions” to predict the dynamic process within the near-surface magma chamber, which is characterized using a pulsed volcanic eruption model.

Moving Target Defenses (MTD) have become a popular and emerging defense strategy for the protection of traditional information technology systems. By their very nature, MTD strategies are designed to protect against adversary reconnaissance efforts on static platforms, essentially sitting back and having unlimited time to identify, craft, execute, and scale an exploit. With the rapid adoption of distributed automotive Cyber-Physical Systems (CPS) ranging from self driving cars, to connected transportation infrastructure, it is becoming more apparent that third party supply chains, increased remote communication interfaces, and legacy software stacks are making the traditionally designed standalone systems become more susceptible to safety-critical cyber-attacks. MTD strategies within the automotive CPS domain have to delicately balance the tradeoff between security and real time predictability, maintaining the safety constraints of the systems. In this paper, we explore the various MTD strategies presented within the literature while discussing potential applicability and strategies sufficient for the automotive CPS domain.