基于ST-ConvLSTM的南海海表面CO₂分压的空间和时间序列预测     

高宇  李爽  郝鹏  宋金宝 《海洋与湖沼》2023,54(6):1573-1585

海表面二氧化碳分压(pCO₂)的未来变化趋势,对统计评估全球碳收支以及理解全球气候变化背景下的海洋酸化现象至关重要。目前传统的海面pCO₂预测方法大部分基于有限的实测数据,然而实测数据存在着时间和地理方面的制约,且计算成本较高。近年来,随着时空观测数据的爆炸性增长,基于深度学习的数据驱动模型在海表面pCO₂预测方面中表现出良好的潜力。然而,由于多种环境因素与海表面pCO₂之间的关系错综复杂,到目前为止尚无十分简单有效的相关模型来对海表面pCO₂进行预测。为应对这一挑战,利用时空卷积长短时记忆神经网络(ST-ConvLSTM)模型,通过海面温度(sea surface temperature, SST)、海面盐度(sea surface salinity, SSS)、叶绿素a浓度(chl a)和海面pCO₂数据,预测南海的海面pCO₂,并将2019年1～12月的数据作为测试集对模型的表现进行了验证。结果显示, ST-ConvLSTM模型...  相似文献   

基于深度学习的倾斜摄影建筑物表面损毁信息提取     

罗嘉琦  帅向华  李继赓 《中国地震》2023,39(2):271-281

建筑物受损信息是地震受灾程度评估的基础,针对传统建筑物表面信息识别人工成本高、效率低等问题,受深度学习提取建筑物影像的启发,提出利用无人机倾斜摄影模型与深度学习相结合的方法提取震后建筑物表面破损信息。以2019年长宁6.0级地震为例,选用双河镇震后倾斜摄影模型切片图为数据源,对比分析面向对象分类方法、VGG-16模型和DeeplabV3+模型对建筑物表面损毁信息的提取结果。分析结果表明,针对建筑物表面破损信息的提取,尤其是细小裂缝的提取,语义分割网络DeeplabV3+模型具有较强的优势(准确率96.93%、召回率96.85%、总体精度96.89%),可实现建筑物表面破损信息的有效提取,具有较强的应用价值。  相似文献   

基于迁移学习的地球物理测井储层参数预测方法研究     

邵蓉波  肖立志  廖广志  周军  李国军 《地球物理学报》2022,65(2):796-808

随着大数据和机器学习的成熟和推广应用,人工神经网络在地球物理测井预测储层参数中得到重视.本文引入迁移学习进行测井储层参数预测,以孔隙度预测神经网络模型和孔隙度含水饱和度联合预测神经网络模型为基础模型,分别以渗透率及含水饱和度预测作为目标任务进行迁移学习,以提升储层参数预测效果和效率.文中详细阐述了基于迁移学习的测井储层...  相似文献   

Combining machine learning techniques,microanalyses and large geochemical datasets for tephrochronological studies in complex volcanic areas: New age constraints for the Pleistocene magmatism of central Italy     

《Quaternary Geochronology》2017

Characterization, correlation and provenance determination of tephra samples in sedimentary sections (tephrochronological studies) are powerful tools for establishing ages of depositional events, volcanic eruptions, and tephra dispersion. Despite the large literature and the advancements in this research field, the univocal attribution of tephra deposits to specific volcanic sources remains too often elusive. In this contribution, we test the application of a machine learning technique named Support Vector Machine to attempt shedding new light upon tephra deposits related to one of the most complex and debated volcanic regions on Earth: the Pliocene-Pleistocene magmatism in Italy. The machine learning algorithm was trained using one of the most comprehensive global petrological databases (GEOROC); 17 chemical elements including major (SiO₂, TiO₂, Al₂O₃, Fe₂O_3T, CaO, MgO, MnO, Na₂O, K₂O, P₂O₅) and selected trace (Sr, Ba, Rb, Zr, Nb, La, Ce) elements were chosen as input parameters. We first show the ability of support vector machines in discriminating among different Pliocene-Pleistocene volcanic provinces in Italy and then apply the same methodology to determine the volcanic source of tephra samples occurring in the Caio outcrop, an Early Pleistocene sedimentary section located in Central Italy. Our results show that: 1) support vector machines can successfully resolve high-dimensional tephrochronological problems overcoming the intrinsic limitation of two- and three-dimensional discrimination diagrams; 2) support vector machines can discriminate among different volcanic provinces in complex magmatic regions; 3) in the specific case study, support vector machines indicate that the most probable source for the investigated tephra samples is the so-called Roman Magmatic Province. These results have strong geochronological and geodynamical implications suggesting new age constraints (1.4 Ma instead of 0.8 Ma) for the starting of the volcanic activity in the Roman Magmatic Province.  相似文献   

地震油气储层的小样本卷积神经网络学习与预测   总被引：2，自引：0，他引：2       下载免费PDF全文

林年添  张栋  张凯  王守进  付超  张建彬  张冲 《地球物理学报》2018,61(10):4110-4125

地震储层预测是油气勘探的重要组成部分,但完成该项工作往往需要经历多个环节,而多工序或长周期的研究分析降低了勘探效率.基于油气藏分布规律及其在地震响应上所具有的特点,本文引入卷积神经网络深度学习方法,用于智能提取、分类并识别地震油气特征.卷积神经网络所具有的强适用性、强泛化能力,使之可以在小样本条件下,对未解释地震数据体进行全局优化提取特征并加以分类,即利用有限的已知含油气井段信息构建卷积核,以地震数据为驱动,借助卷积神经网络提取、识别蕴藏其中的地震油气特征.将本方案应用于模型数据及实际数据的验算,取得了预期效果.通过与实际钻井信息及基于多波地震数据机器学习所预测结果对比,本方案利用实际数据所演算结果与实际情况有较高的吻合度.表明本方案具有一定的可行性,为缩短相关环节的周期提供了一种新的途径.  相似文献   

A model‐based data‐driven dictionary learning for seismic data representation          下载免费PDF全文

Can Evren Yarman  Rajiv Kumar  James Rickett 《Geophysical Prospecting》2018,66(1):98-123

Planar waves events recorded in a seismic array can be represented as lines in the Fourier domain. However, in the real world, seismic events usually have curvature or amplitude variability, which means that their Fourier transforms are no longer strictly linear but rather occupy conic regions of the Fourier domain that are narrow at low frequencies but broaden at high frequencies where the effect of curvature becomes more pronounced. One can consider these regions as localised “signal cones”. In this work, we consider a space–time variable signal cone to model the seismic data. The variability of the signal cone is obtained through scaling, slanting, and translation of the kernel for cone‐limited (C‐limited) functions (functions whose Fourier transform lives within a cone) or C‐Gaussian function (a multivariate function whose Fourier transform decays exponentially with respect to slowness and frequency), which constitutes our dictionary. We find a discrete number of scaling, slanting, and translation parameters from a continuum by optimally matching the data. This is a non‐linear optimisation problem, which we address by a fixed‐point method that utilises a variable projection method with ?₁ constraints on the linear parameters and bound constraints on the non‐linear parameters. We observe that slow decay and oscillatory behaviour of the kernel for C‐limited functions constitute bottlenecks for the optimisation problem, which we partially overcome by the C‐Gaussian function. We demonstrate our method through an interpolation example. We present the interpolation result using the estimated parameters obtained from the proposed method and compare it with those obtained using sparsity‐promoting curvelet decomposition, matching pursuit Fourier interpolation, and sparsity‐promoting plane‐wave decomposition methods.  相似文献   

基于深度学习的赤潮发生预报方法研究     

余璇  石绥祥  徐凌宇  王蕾 《海洋通报》2021,40(5)

赤潮作为海洋灾害,对海洋渔业、生态、经济,以及人类生产、生活造成了严重影响。一直以来,赤潮受到研究者的广泛关注,但由于它的形成机制比较复杂,使得赤潮预报极具挑战性。针对赤潮预报的研究问题,本文收集了厦门海域赤潮发生前后的海洋监测数据,结合皮尔逊相关系数、散布矩阵、复相关系数方法,分析多环境因子与赤潮发生多要素的关联情况,重点采用基于深度学习的LSTM与CNN融合方法,挖掘环境因子的时序依赖,发现序列数据的局部特征,对赤潮发生进行预报。在厦门一号和厦门二号数据集中,本方法在预报未来12 h内的赤潮情况时,RMSE、MAE误差分别达到0.521 8、0.504 3。通过协同对比模型进一步确定赤潮发生的预报概率,在两个数据集上的最终预报准确率分别为67.58%和63.49%。本研究为赤潮的分析预报提供了探索经验,证明了将深度学习方法应用于赤潮预报的可行性。  相似文献   

基于LBS和深度学习的旅游景区客流量的高时频预测     

谢谦  陆明  谢春山 《地球信息科学学报》2023,25(2):298-310

为实现精准的旅游景区客流量的高时频预测,本研究构建了一套基于LBS和深度学习模型的预测方法。此方法可通过对LBS数据的转换实现预测的空间范围与时频控制,并通过方法的核心模型——基于双向循环神经网络和GRU算法构建的深度双向GRU(DBi-GRU)模型完成预测。为检验方法的有效性,研究以深圳大梅沙海滨公园为例对方法进行实验测试。实验使用拟合曲线、误差指标及DM检验3种方法评估DBi-GRU模型的预测效果。此外,实验还设置了其他五种深度学习模型作为DBi-GRU的对照模型,测试基于不同深度学习算法的模型之间的预测水平差异。实验结果表明:(1)本研究提出的DBi-GRU模型在景区客流量高时频预测中具有理想的预测效果,在高峰时段的客流量预测方面也具有较高准确性,预测效果明显优于其他深度学习模型;(2)基于双向循环网络的模型的效果普遍优于基于常规循环网络的模型。尤其是基于双向LSTM算法的模型,虽然预测的准确度略逊色于DBi-GRU模型,但在模型性能上与其的差异并不显著;(3)在相同网络参数下,GRU算法较前人采用的LSTM和RNN算法有着更高的预测准确性。本研究为客流量预测领域的研究提供了一种...  相似文献   

Data-based modelling approach for variable density flow and solute transport simulation in a coastal aquifer   总被引：1，自引：1，他引：0  

Basant Yadav  Shashi Mathur  Sudheer Ch  Brijesh Kumar Yadav 《水文科学杂志》2018,63(2):210-226

Data-based models, namely artificial neural network (ANN), support vector machine (SVM), genetic programming (GP) and extreme learning machine (ELM), were developed to approximate three-dimensional, density-dependent flow and transport processes in a coastal aquifer. A simulation model, SEAWAT, was used to generate data required for the training and testing of the data-based models. Statistical analysis of the simulation results obtained by the four models show that the data-based models could simulate the complex salt water intrusion process successfully. The selected models were also compared based on their computational ability, and the results show that the ELM is the fastest technique, taking just 0.5 s to simulate the dataset; however, the SVM is the most accurate, with a Nash-Sutcliffe efficiency (NSE) ≥ 0.95 and correlation coefficient R ≥ 0.92 for all the wells. The root mean square error (RMSE) for the SVM is also significantly less, ranging from 12.28 to 77.61 mg/L.  相似文献   

多尺度特征融合深度学习建筑物的提取方法     

刘恒恒  张春森  葛英伟  史书 《地理空间信息》2022,20(2):97-100

提出一种基于多尺度特征融合的建筑物提取方法,结合新的网络DenseASPP-UNet,以实现影像多尺度特征的融合,进而高精度提取建筑物.通过Inria开源建筑物航空影像数据集进行验证,表明DenseASPP-UNet相比其他深度学习方法建筑物提取精度有很大的提升.  相似文献