基于街景图像的可解释性城市感知模型研究方法

理解城市环境对人类感知的影响,对城市合理规划及布局具有重要的人文参考价值。城市环境是一个动态变化的复杂系统,具有空间异质性的特点。由于研究方法的限制,在复杂的城市环境中,以往基于街景图像的城市感知研究难以全面精细地分析环境关键要素对人类感知的影响。本研究以武汉市中心为研究区,首先利用全卷积神经网络将街景图像分割为城市地物类型,耦合感知打分数据和随机森林算法建立6类城市感知模型;然后基于沙普利值方法分解在随机森林模型中各类城市地物对人类感知的影响,并识别城市环境关键要素;最后结合分解结果,探究在非线性模型中沙普利值方法的适用性和优势。结果表明：沙普利值方法能够有效考虑环境异质性,精确地定量表示在不同场景中各类地物对人类感知的影响;城市高楼、天空、绿地空间是对人类感知影响最大的3类地物,且地物的体积和分布与其对人类感知的影响有关,图像占比大、分布连续的地物对人类感知的影响比图像占比小、分布离散的地物对人类感知的影响大;受城市环境空间异质性的影响,主要地物类型对各类感知的影响程度和形式有显著不同;高楼与人类感知为非线性关系,且具有明显的单调递增或递减的形式;绿地空间与积极感知呈非线性关系,与...     

Geophysical mapping of hyporheic processes controlled by sedimentary architecture within compound bar deposits

Hyporheic exchange influences water quality and controls numerous physical, chemical, and biological processes. Despite its importance, hyporheic exchange and the associated dynamics of solute mixing are often difficult to characterize due to spatial (e.g., sedimentary heterogeneity) and temporal (e.g., river stage fluctuation) variabilities. This study coupled geophysical techniques with physical and chemical sediment analyses to map sedimentary architecture and quantify its influence on hyporheic exchange dynamics within a compound bar deposit in a gravel-dominated river system in southwestern Ohio. Electromagnetic induction (EMI) was used to quantify variability in electrical conductivity within the compound bar. EMI informed locations of electrode placement for time-lapse electrical resistivity imaging (ERI) surveys, which were used to examine changes in electrical resistivity driven by hyporheic exchange. Both geophysical methods revealed a zone of high electrical conductivity in the center of the bar, identified as a fine-grained cross-bar channel fill. The zone acts as a baffle to flow, evidenced by stable electrical conditions measured by time-lapse ERI over the study period. Large changes in electrical resistivity throughout the survey period indicate preferential flowpaths through higher permeability sands and gravels. Grain size analyses confirmed sedimentological interpretations of geophysical data. Loss on ignition and x-ray fluorescence identified zones with higher organic matter content that are locations for potentially enhanced geochemical activity within the cross-bar channel fill. Differences in the physical and geochemical characteristics of cross-bar channel fills play an important role in hyporheic flow dynamics and nutrient processing within riverbed sediments. These findings enhance our understanding of the applications of geophysical methods in mapping riverbed heterogeneity and highlight the importance of accurately representing geomorphologic features and heterogeneity when studying hyporheic exchange processes.     

Near-surface water fluxes and their controls in a sloping heterogeneously layered volcanic soil beneath a supra-wet tropical montane cloud forest (NW Costa Rica)

Tropical montane cloud forests (TMCF) receive additional (‘occult’) inputs of water from fog and wind-driven rain. Together with the concomitant reduction in evaporative losses, this typically leads to high soil moisture levels (often approaching saturation) that are likely to promote rapid subsurface flow via macropores. Although TMCF make up an estimated 6.6% of all remaining montane tropical forest and occur mostly in steep headwater areas that are protected in the expectation of reduced downstream flooding, TMCF hillslope hydrological functioning has rarely been studied. To better understand the hydrological response of a supra-wet TMCF (net precipitation up to 6535 mm y⁻¹) on heterogeneously layered volcanic ash soils (Andosols), we examined temporal and spatial soil moisture dynamics and their contribution to shallow subsurface runoff and stormflow for a year (1 July 2003–30 June 2004) in a small headwater catchment on the Atlantic (windward) slope near Monteverde, NW Costa Rica. Particular attention was paid to the partitioning of water fluxes into lateral subsurface flow and vertical percolation. The presence of a gravelly layer (C-horizon) at ~25 cm depth of very high hydraulic conductivity (geometric mean: 502 mm h⁻¹) intercalated between two layers of much lower conductivity (7.5 and 15.7 mm h⁻¹ above and below, respectively), controlled both surface infiltration and delayed vertical water movement deeper into the soil profile. Soil water fluxes during rainfall were dominated by rapid lateral flow in the gravelly layer, particularly at high soil moisture levels. In turn, this lateral subsurface flow controlled the magnitude and timing of stormflow from the catchment. Stormflow amount increased rapidly once topsoil moisture content exceeded a threshold value of ~0.58 cm³ cm⁻³. Responses were not affected appreciably by rainfall intensity because soil hydraulic conductivities across the profile largely exceeded prevailing rainfall intensities.     

Artificial intelligence for identifying hydrologically homogeneous regions: A state‐of‐the‐art regional flood frequency analysis

Due to the severity related to extreme flood events, recent efforts have focused on the development of reliable methods for design flood estimation. Historical streamflow series correspond to the most reliable information source for such estimation; however, they have temporal and spatial limitations that may be minimized by means of regional flood frequency analysis (RFFA). Several studies have emphasized that the identification of hydrologically homogeneous regions is the most important and challenging step in an RFFA. This study aims to identify state‐of‐the‐art clustering techniques (e.g., K ‐means, partition around medoids, fuzzy C‐means, K ‐harmonic means, and genetic K ‐means) with potential to form hydrologically homogeneous regions for flood regionalization in Southern Brazil. The applicability of some probability density function, such as generalized extreme value, generalized logistic, generalized normal, and Pearson type 3, was evaluated based on the regions formed. Among all the 15 possible combinations of the aforementioned clustering techniques and the Euclidian, Mahalanobis, and Manhattan distance measures, the five best were selected. Several watersheds' physiographic and climatological attributes were chosen to derive multiple regression equations for all the combinations. The accuracy of the equations was quantified with respect to adjusted coefficient of determination, root mean square error, and Nash–Sutcliffe coefficient, whereas, a cross‐validation procedure was applied to check their reliability. It was concluded that reliable results were obtained when using robust clustering techniques based on fuzzy logic (e.g., K ‐harmonic means), which have not been commonly used in RFFA. Furthermore, the probability density functions were capable of representing the regional annual maximum streamflows. Drainage area, main river length, and mean altitude of the watershed were the most recurrent attributes for modelling of mean annual maximum streamflow. Finally, an integration of all the five best combinations stands out as a robust, reliable, and simple tool for estimation of design floods.     

Numerical modeling of fracturing in permeable rocks via a micromechanical continuum model

The paper presents a micromechanical approach to describe the failure of low-permeability brittle rocks as a multiscale fracturing process based on a poroelastic microcrack-damage model. Failure is formulated deep down at the fine pore scale as a material degradation phenomenon driven by microcrack growth that also impacts upon hydromechanical properties. A set of damage tensors describes the effect of dual-scale porosities (nanopores and microcracks) on both the hydraulic and poroelastic rock properties. Essentially, the multiscale model reconstructs the coupling effect of hydromechanical forces at the continuum level from the ground up through the upscaling of multiphase interactions at the fundamental structural level of the material. As a result, many macroscopic characteristics emerge naturally such as friction angle, fracture properties, and most importantly, Biot's coefficient taking on a tensorial form that is generally anisotropic. The model is validated within the framework of finite elements to illustrate various baseline constitutive features such as the effect of microcrack growth on the nonlinear stress-strain response and the induced anisotropy in the context of lab experimental tests and boundary value problems. Heterogeneities of the rock samples were incorporated by choosing material properties to be stochastic following Weibull and lognormal distributions. Numerical results appropriately replicated typical experimental observations where fracture localization and propagation are shown to be a multiscale phenomenon emerging from microcrack growth and coalescence at the microscale, with concomitant enhancement in fluid conductivity.     

Surface water-groundwater exchange dynamics in buried-valley aquifer systems

Groundwater is a primary source of drinking water worldwide, but excess nutrients and emerging contaminants could compromise groundwater quality and limit its usage as a drinking water source. As such contaminants become increasingly prevalent in the biosphere, a fundamental understanding of their fate and transport in groundwater systems is necessary to implement successful remediation strategies. The dynamics of surface water-groundwater (hyporheic) exchange within a glacial, buried-valley aquifer system are examined in the context of their implications for the transport of nutrients and contaminants in riparian sediments. High conductivity facies act as preferential flow pathways which enhance nutrient and contaminant delivery, especially during storm events, but transport throughout the aquifer also depends on subsurface sedimentary architecture (e.g. interbedded high and low conductivity facies). Temperature and specific conductance measurements indicate extensive hyporheic mixing close to the river channel, but surface water influence was also observed far from the stream-aquifer interface. Measurements of river stage and hydraulic head indicate that significant flows during storms (i.e., hot moments) alter groundwater flow patterns, even between consecutive storm events, as riverbed conductivity and, more importantly, the hydraulic connectivity between the river and aquifer change. Given the similar mass transport characteristics among buried-valley aquifers, these findings are likely representative of glacial aquifer systems worldwide. Our results suggest that water resources management decisions based on average (base) flow conditions may inaccurately represent the system being evaluated, and could reduce the effectiveness of remediation strategies for nutrients and emerging contaminants.     

Anisotropy of heterogeneity scale lengths in the lower mantle from PKIKP precursors

The anisotropy of heterogeneity scale lengths in the lower mantle is investigated by modelling its effect on the high-frequency precursors of PKIKP scattered by the heterogeneities. Although models having either an isotropic or an anisotropic distribution of scale lengths can fit the observed coda shapes of short-period precursors, the frequency content of broad-band PKIKP precursors favours a dominantly isotropic distribution of scale lengths. Precursor coda shapes are consistent with 1 per cent fluctuations in P velocity in the wavenumber band 0.05–0.5  km⁻¹ extending to 1000  km above the core–mantle boundary, and with a D" region open to circulation throughout the lower mantle. The level of excitation of PKIKP precursors observed in the frequency band 0.02–2  Hz requires a power spectrum of heterogeneity that is nearly white or slowly increasing with wavenumber. Anisotropy of scale lengths may exist in a D" layer having larger horizontal than vertical scale lengths and produce little or no detectable effects on PKIKP precursors for P -velocity perturbations as high as 3 per cent when averaged over a vertical scale of several kilometres, and much higher when averaged over scales of hundreds of metres or less.     

利用地震波场参数探讨岩性非均质的分布

本文利用 Ｓ Ｔ Ｍ 地区二维地震资料中提取中油组的主振幅参数与岩性参数进行相关拟合,并依据岩性分析资料和调谐厚度将其划分为三个区带：低振幅带、正常振幅带和异常高振幅带。综合分析油气测试结果、岩性统计资料后认为,其中的异常高振幅带是结构疏松、孔隙发育、连通性好的均质储层分布区,并从合成记录和理论模型正演结果对上述结论进行了证实。结合沉积相分析成果,将该地区的岩性非均质划分为三种类型,后期在优质储层均质区块中钻探获得了高产工业油气流,从而证实了从地震角度划分岩性非均质分布,不仅是可行的,而且是准确的。该方法不仅拓宽了地震资料的应用范围,也为寻找有利油气富集区提供了一种行之有效的方法。     

辽宁复县金刚石的阴极发光特征及其意义

阴极发光图像揭示了辽宁复县金刚石的内部结构具有３种类型;（１）简单的生长环带结构;（２）多期生长的复杂环带结构;（３）稀少的似玛瑙状结构,这些内部结构特征反映金刚石的不均一性及生长具多期、多阶段特点、该特点与金刚石所处的流体－熔体的不同阶段相对应。玛瑙状结构除多生长中心聚集形成复杂种晶形态所致外,局部不均匀熔蚀、混合的生长机制的变化的结晶条件三者综合作用亦是原因之一,阴有发光图像还反映了金刚石晶体