基于机器学习的海洋环境预报订正方法研究     

许立兵  王安喜  汪纯阳  陈悦  陈昱文  周峥  陈幸荣  邢建勇  刘克威  黄小猛 《海洋通报》2020,39(6):695-704

结合中尺度数值模式 WRF 预报数据和 ERA5 再分析资料，利用机器学习方法对 WRF 预报场的风场、温度、气压进行预报订正。采用 ERA5 作为真值，与原始 WRF 预报相比，利用随机森林模型可以将预报结果整体均方根误差降低 44%以 上，利用深度神经网络模型可以将预报结果整体均方根误差降低 34%以上。通过随机森林模型实验得到不同输入特征对预报要素的影响程度，分析了关键的预报订正因子。  相似文献   

Simulation of site-scale water fluxes in desert and natural oasis ecosystems of the arid region in Northwest China     

Mingjuan Xie  Geping Luo  Olaf Hellwich  Amaury Frankl  Wenqiang Zhang  Chunbo Chen  Chen Zhang  Philippe De Maeyer 《水文研究》2021,35(12):e14444

The study of water fluxes is important to better understand hydrological cycles in arid regions. Data-driven machine learning models have been recently applied to water flux simulation. Previous studies have built site-scale simulation models of water fluxes for individual sites separately, requiring a large amount of data from each site and significant computation time. For arid areas, there is no consensus as to the optimal model and variable selection method to simulate water fluxes. Using data from seven flux observation sites in the arid region of Northwest China, this study compared the performance of random forest (RF), support vector machine (SVM), back propagation neural network (BPNN), and multiple linear regression (MLR) models in simulating water fluxes. Additionally, the study investigated inter-annual and seasonal variation in water fluxes and the dominant drivers of this variation at different sites. A universal simulation model for water flux was built using the RF approach and key variables as determined by MLR, incorporating data from all sites. Model performance of the SVM algorithm (R² = 0.25–0.90) was slightly worse than that of the RF algorithm (R² = 0.41–0.91); the BPNN algorithm performed poorly in most cases (R² = 0.15–0.88). Similarly, the MLR results were limited and unreliable (R² = 0.00–0.66). Using the universal RF model, annual water fluxes were found to be much higher than the precipitation received at each site, and natural oases showed higher fluxes than desert ecosystems. Water fluxes were highest during the growing season (May–September) and lowest during the non-growing season (October–April). Furthermore, the dominant drivers of water flux variation were various among different sites, but the normalized difference vegetation index (NDVI), soil moisture and soil temperature were important at most sites. This study provides useful insights for simulating water fluxes in desert and oasis ecosystems, understanding patterns of variation and the underlying mechanisms. Besides, these results can make a contribution as the decision-making basis to the water management in desert and oasis ecosystems.  相似文献   

Site characterization model using least‐square support vector machine and relevance vector machine based on corrected SPT data (Nc)     

Pijush Samui  T. G. Sitharam 《国际地质力学数值与分析法杂志》2010,34(7):755-770

Statistical learning algorithms provide a viable framework for geotechnical engineering modeling. This paper describes two statistical learning algorithms applied for site characterization modeling based on standard penetration test (SPT) data. More than 2700 field SPT values (N) have been collected from 766 boreholes spread over an area of 220 sqkm area in Bangalore. To get N corrected value (N_c), N values have been corrected (N_c) for different parameters such as overburden stress, size of borehole, type of sampler, length of connecting rod, etc. In three‐dimensional site characterization model, the function N_c=N_c (X, Y, Z), where X, Y and Z are the coordinates of a point corresponding to N_c value, is to be approximated in which N_c value at any half‐space point in Bangalore can be determined. The first algorithm uses least‐square support vector machine (LSSVM), which is related to a ridge regression type of support vector machine. The second algorithm uses relevance vector machine (RVM), which combines the strengths of kernel‐based methods and Bayesian theory to establish the relationships between a set of input vectors and a desired output. The paper also presents the comparative study between the developed LSSVM and RVM model for site characterization. Copyright © 2009 John Wiley & Sons, Ltd.  相似文献   

测斜技术原理与作业要点探讨     

侯金波  李更召  陈潇  朱俊鹏 《测绘与空间地理信息》2016,(6)

深层水平位移监测(即测斜)能直接反映土体或支护结构的变形特征,是高风险建筑基坑工程安全的重要保证。本文对测斜技术原理进行了阐述,并根据工程实践经验对测斜监测的作业要点进行了探讨与总结。  相似文献   

迁移学习支持下的高分影像山地滑坡灾害解译模型     

郭加伟  李永树  李政  刘锟铭  张帅毅 《测绘科学技术学报》2016,(5):496-501

无人机低空遥感是近年来新兴的一种快速获取灾情信息的手段,如何利用无人机高分影像构建滑坡灾害解译模型是实现快速自动解译滑坡的关键。针对该问题,对比了多种影像特征提取方法,将迁移学习(TL)特征和支持向量机(SVM)引入到构建滑坡灾害自动解译模型中,提出了一种TL支持下的高分影像滑坡灾害解译模型。选取5·12汶川地震及4·20芦山地震系列无人机影像构建了滑坡灾害样本库并进行了实验,TL特征方法整体分类准确度ACC为95%,ROC达到0.98,识别准确率达到97%。结果表明,所提方法可用于高分影像滑坡自动解译,同时可用于大面积高分影像中快速山地滑坡灾害定位及检测。  相似文献   

Flow processes and sedimentation in unidirectionally migrating deep‐water channels: From a three‐dimensional seismic perspective          下载免费PDF全文

Chenglin Gong  Yingmin Wang  Ronald J. Steel  Jeff Peakall  Xiaoming Zhao  Qiliang Sun 《Sedimentology》2016,63(3):645-661

Three‐dimensional seismic data were used to infer how bottom currents control unidirectional channel migration. Bottom currents flowing towards the steep bank would deflect the upper part of sediment gravity flows at an orientation of 1° to 11° to the steep bank, yielding a helical flow circulation consisting of a faster near‐surface flow towards the steep bank and a slower basal return flow towards the gentle bank. This helical flow model is evidenced by the occurrence of bigger, muddier (suggested by low‐amplitude seismic reflections) lateral accretion deposits and gentle channel wall with downlap terminations on the gentle bank and by smaller, sandier (indicated by high‐amplitude seismic reflectors) channel fills and steep channel walls with truncation terminations on the steep bank. This helical flow circulation promotes asymmetrical depositional patterns with dipping accretion sets restricted to the gentle bank, which restricts the development of sinuosity and yields unidirectional channel migration. These results aid in obtaining a complete picture of flow processes and sedimentation in submarine channels.  相似文献   

豫西洛宁县铁炉坪Ag多金属矿床的CSAMT剖面及找矿意义     

梁涛 《地质与勘探》2016,52(2):239-245

为获得新的深部找矿靶区,在铁炉坪Ag多金属矿床实施了CSAMT法测深。CSAMT剖面中,高、中、低电阻率异常体边界清晰、形态各异,高阻异常体具有上、下双层结构,低阻异常体出露较小,多呈孤立形态产出,主要有400m低阻异常体和1400m低阻异常体。高阻异常体被解释为浅部和深部两层岩浆和深部流体活动的产物,推测深部高阻异常体以岩浆体系为主,浅部高阻异常体应是岩浆体系、深部流体体系及围岩硅化蚀变等的综合反映。400m低阻异常体对应于于地表剖面线400m~600m之间的主矿脉,表明它是深部成矿流体的显示,提出1400m低阻异常体为新的深部找矿靶区。  相似文献   

Weak ENSO Asymmetry Due to Weak Nonlinear Air–Sea Interaction in CMIP5 Climate Models   总被引：1，自引：0，他引：1  

Yan Sun  Fan Wang  De-Zheng Sun 《大气科学进展》2016,33(3):352-364

State-of-the-art climate models have long-standing intrinsic biases that limit their simulation and projection capabilities.Significantly weak ENSO asymmetry and weakly nonlinear air–sea interaction over the tropical Pacific was found in CMIP5(Coupled Model Intercomparison Project, Phase 5) climate models compared with observation. The results suggest that a weak nonlinear air–sea interaction may play a role in the weak ENSO asymmetry. Moreover, a weak nonlinearity in air–sea interaction in the models may be associated with the biases in the mean climate—the cold biases in the equatorial central Pacific. The excessive cold tongue bias pushes the deep convection far west to the western Pacific warm pool region and suppresses its development in the central equatorial Pacific. The deep convection has difficulties in further moving to the eastern equatorial Pacific, especially during extreme El Ni o events, which confines the westerly wind anomaly to the western Pacific. This weakens the eastern Pacific El Ni o events, especially the extreme El Ni o events, and thus leads to the weakened ENSO asymmetry in climate models. An accurate mean state structure(especially a realistic cold tongue and deep convection) is critical to reproducing ENSO events in climate models. Our evaluation also revealed that ENSO statistics in CMIP5 climate models are slightly improved compared with those of CMIP3. The weak ENSO asymmetry in CMIP5 is closer to the observation. It is more evident in CMIP5 that strong ENSO activities are usually accompanied by strong ENSO asymmetry, and the diversity of ENSO amplitude is reduced.  相似文献   

利用多光谱卫星遥感和深度学习方法进行青藏高原积雪判识     

阚希  张永宏  曹庭  王剑庚  田伟 《测绘学报》2016,45(10):1210-1221

青藏高原积雪对全球气候变化十分重要,针对已有积雪遥感判识方法中普遍采用的可见光与红外光谱数据易受复杂地形与高海拔影响,导致青藏高原地区积雪判识精度较低的问题,提出了一种基于多光谱遥感与地理信息数据特征级融合的积雪遥感判识方法:以风云三号卫星可见光与红外多光谱遥感资料与多要素地理信息作为数据源,由地面实测雪深数据与现有积雪产品交叉筛选出样本标签,构建并训练基于层叠去噪自编码器(SDAE)的特征融合与分类网络,从而有效辨识青藏高原遥感图像中的云、积雪以及无雪地表。经地面实测雪深数据验证,该方法分类精度显著高于使用相同数据源的FY-3A/MULSS积雪产品,略高于国际主流积雪产品MOD10A1与MYD10A1,并且年均云覆盖率最低。试验结果表明该方法可有效地减少云层对积雪判识的干扰,提升分类精度。  相似文献   

历史数据和强化学习相结合的低频轨迹数据匹配算法     

孙文彬  熊婷 《测绘学报》2016,45(11):1328-1334

针对低频(采样间隔大于1min)轨迹数据匹配算法精度不高的问题,提出了一种基于强化学习和历史轨迹的匹配算法HMDP-Q,首先通过增量匹配算法提取历史路径作为历史参考经验库;根据历史参考经验库、最短路径和可达性筛选候选路径集;再将地图匹配过程建模成马尔科夫决策过程,利用轨迹点偏离道路距离和历史轨迹构建回报函数;然后借助强化学习算法求解马尔科夫决策过程的最大回报值,即轨迹与道路的最优匹配结果;最后应用某市浮动车轨迹数据进行试验。结果表明:本文算法能有效提高轨迹数据与道路匹配精度;本算法在1min低频采样间隔下轨迹匹配准确率达到了89.2%;采样频率为16min时,该算法匹配精度也能达到61.4%;与IVVM算法相比,HMDP-Q算法匹配精度和求解效率均优于IVVM算法,16min采样频率时本文算法轨迹匹配精度提高了26%。  相似文献