基于傅里叶变换的频谱分析法在X射线轮胎检测中的应用

子午线轮胎的缺陷检测方法比较多,频谱分析法是基于频域的图像分析方法,其主要功能是获取图像在频域中的纹理特征。傅里叶变换能够将时域的信号分解为频域各频率分量下之和,即分解为各个频率下的谱。然后通过对得到的频谱进行图像处理,判断轮胎有无缺陷。实验证明使用基于傅里叶变换的频谱分析法进行子午线轮胎缺陷检测可以大幅降低工作量,提高轮胎检测效率。     

湘鄂西地区五峰—龙马溪组页岩地层与沉积特征

以湘鄂西及周缘地区晚奥陶世至早志留世沉积的泥页岩为研究对象,本文通过钻井取心测井、野外观察和分析化验等资料,分析了研究区五峰-龙马溪组页岩的发育和沉积特征.结果表明,湘鄂西地区页岩普遍厚40-60 m,纵向上可划分为三亚段,九小层,其中1亚段对应底部①-⑤小层,为高伽马值(GR),低密度,高有机质丰度的页岩段,以灰黑色—黑色硅质-炭质页岩为主,湘鄂西古隆起及周缘①-⑤小层页岩段有不同程度的缺失.页岩纵向上发育深水硅质陆棚、深水砂质陆棚、深水混积陆棚和深水泥质陆棚4种沉积微相,其中①-⑤小层页岩段对应深水硅质陆棚微相,为缺氧的强还原沉积环境,沉积了富含笔石、黄铁矿的富硅、富碳页岩,其硅质多为自生成因,对应焦石坝目前的主力产气页岩段.     

三峡库区树坪滑坡地下水渗流场分析

库岸边坡的地下水分布对其稳定性影响重大。以三峡库区树坪滑坡为例,基于Visual Modflow软件,对不同工况下的地下水渗流场进行数值模拟。结论表明：滑坡地下水位随着库水位波动而变化; 滑坡体后缘地下水受到影响较小,中部和前缘水头变化明显; 在库水位下降过程中,滑坡体内地下水随时间变化呈现水位不变、加速下降、降速减缓、稳定水位的趋势,且不同库水降速只对地下水位变化速率产生影响,对最终地下水位没有影响; 库水位降速越大,滑坡前缘地下水位变化越强烈,水力梯度也越大,且在滑坡东西两侧边界处形成的水力梯度最大。其成果将为库岸滑坡渗流场变化及稳定性分析提供一定的参考价值。     

内蒙古阿巴嘎旗锡力诺尔金矿地质特征及成矿作用探讨

锡力诺尔金矿位于二连-贺根山断裂带偏南侧,为国土资源大调查中新发现的金矿点。本文在系统研究苏左旗-阿巴嘎旗地区成矿地质背景和成矿规律基础上,对锡力诺尔金矿点的地质特征和控矿因素进行了阐述,认为金矿化体主要赋存于二叠系上统林西组的细砂岩、砂砾岩的层间破碎带中,受控于北东向韧性剪切构造带和次级张性裂隙。通过锡力诺尔金矿点与外围金矿点的对比分析,初步认为锡力诺尔金矿体的成因类型为石英脉与蚀变岩复合型,矿体主体为低品位蚀变岩型,显示出较好的找矿前景。本文可为摸清该地区的金矿找矿思路和圈定金矿找矿靶区提供借鉴。     

上海青浦CINRAD/SAD雷达和南汇WSR-88D雷达数据对比分析

上海青浦CINRAD/SAD雷达和南汇WSR 88D雷达型号不同,并且采用不同的双偏振技术升级方案,实际业务运行中2部雷达存在一定的数据偏差。为了对雷达的探测性能进行定量评估,在共同探测区域遴选2022年7—9月及2023年6月共30个过程3933个时次数据,以稳定性降水与对流性降水分类并按强度分级进行对比,评估了反射率因子ZH,差分反射率ZDR,相关系数CC和差分传播相移ΦDP的水平分布特征及随方位变化趋势,雷达灵敏度随仰角变化趋势以及ZH和ZDR偏差的定量统计。结果表明：2部雷达ZH的观测结果接近,青浦CINRAD/SAD雷达的ZH,CC和ΦDP受地物及避雷针影响较大,随方位变化不稳定,青浦CINRAD/SAD雷达在低层的非气象回波滤除效果较南汇WSR 88D雷达差。2部雷达ZH和ZDR的偏差在ZH大于40 dBz的区域较大,ZH平均偏差为0.9 dB,ZDR为-0.14 dB。本文基于台风、冰雹、短时强降水和梅雨降水数据,定量化评估了上海2部S波段双偏振天气雷达数据之间的差异,为后续算法改进以及数据订正提供参考。     

冀南綦村斑状石英二长岩角闪石环带结构及其地质意义

为阐释岩浆两阶段演化过程及其与铁成矿的关系,对邯邢铁矿区斑状石英二长岩中角闪石斑晶环带结构进行了成因矿物学研究.通过角闪石温度计、压力计、氧逸度计和湿度计的计算,发现岩浆在14km深处岩浆房内经过多次岩浆混合过程,其后演化的岩浆侵位到1.9 km的深度.岩浆房的温度为935～964℃,氧逸度为△NNO+0.27～△NN...     

高精度磁测在IOCG型铁矿勘查中的应用——以智利英格瓦塞铁矿为例

IOCG型（铁氧化物铜金型）矿床富含铁氧化物,且常缺失硫化物,高精度磁测是寻找IOCG矿床重要的有效手段之一。英格瓦塞铁矿位于智利中北部IOCG型矿床成矿带,前期地质工作程度较低。在智利英格瓦塞铁矿矿区地质特征调查基础上,重点针对中部矿区的7个磁异常区,利用RGIS对其高精度磁测数据进行面积上延、面积下延、剖面下延及2.5D反演拟合处理,结果表明,本区引起磁异常的磁性地质体整体呈NE向延伸、倾向NW、倾角较陡。除Ⅰ号和Ⅴ号磁异常规模较小外,其他几个磁异常中心找矿前景较大,有规模的磁异常区大多埋深在50~120 m之间,为下一步地质找矿钻探验证提供了有效的基础。同时,该勘查方法可为中国IOCG型铁矿找矿工作提供参考依据。     

Temporal changes in wetland plant communities with decades of cumulative water pollution in two plateau lakes in China’s Yunnan Province

Wetland plant communities in the plateau lakes of Yunnan Province, China, have decreased significantly over the past decades. To better understand this degradation, we analyzed the processes and characteristics of changes in wetland plant communities in two of the largest lakes in Yunnan Province, Dianchi and Erhai lakes. We collected records of native and alien plant communities in the two lakes from literature published from the 1950s to current period. We calculated plant community types and their area in some historical periods when related data were reported, and analyzed the relationship between changes in plant communities and water pollution. In Dianchi Lake, 12 community types of native plant communities, covering over 80% of the surface in the 1950s and 1960s, were reduced to four types covering 2.4% by the late 2000s. Alien plant communities started to appear in the lake in the late 1970s, and have since come to cover 4.9% of the lake surface, thereby becoming dominant. In Erhai Lake, 16 community types of native plant communities, covering 47.1% of the lake surface in the late 1970s, declined to 10 community types, covering 9.3% of the surface, by the late 2000s. Alien plant communities appeared in the middle 1980s, and at present cover 0.7% of the surface area. It was indicated that changes in plant communities were significant related to water eutrophication. The area occupied by native and alien plant communities was, respectively, negatively and positively related to the content of nutrients in water. This showed lacustrine pollution played an important role in native plant loss and alien plant invasion in the two plateau lakes.     

Limit equilibrium method (LEM) of slope stability and calculation of comprehensive factor of safety with double strength-reduction technique

When the slope is in critical limit equilibrium(LE) state, the strength parameters have different contribution to each other on maintaining slope stability. That is to say that the strength parameters are not simultaneously reduced. Hence, the LE stress method is established to analyze the slope stability by employing the double strengthreduction(DSR) technique in this work. For calculation model of slope stability under the DSR technique, the general nonlinear Mohr–Coulomb(M–C) criterion is used to describe the shear failure of slope. Meanwhile, the average and polar diameter methods via the DSR technique are both adopted to calculate the comprehensive factor of safety(FOS) of slope. To extend the application of the polar diameter method, the original method is improved in the proposed method. After comparison and analysis on some slope examples, the proposed method's feasibility is verified. Thereafter, the stability charts of slope suitable for engineering application are drawn. Moreover, the studies show that:(1) the average method yields similar results as that of the polardiameter method;(2) compared with the traditional uniform strength-reduction(USR) technique, the slope stability obtained using the DSR techniquetends to be more unsafe; and(3) for a slope in the critical LE state, the strength parameter φ, i.e., internal friction angle, has greater contribution on the slope stability than the strength parameters c, i.e., cohesion.     

基于贝叶斯机器学习的生态模型参数优化方法研究

参数优化方法是准确估计生态模型参数、降低其不确定性的有效手段。本文提出一种基于贝叶斯机器学习的No-U-Turn Sampler（NUTS）生态模型参数优化方法。NUTS是一种高效的参数优化方法,每次取样中利用递归算法生成候选参数集（二叉树）推断参数的后验信息,如果满足约束条件“非U型回转”,不断构建子树更新参数;否则,记录本次抽样的“最优”参数集,并开始下一次取样,直到获取足够样本。该算法在每次取样中充分优化参数,避免因随机游走行为产生冗余抽样,提高了参数优化效率。本文以千烟洲亚热带人工针叶林碳通量模拟为例,基于Pymc3框架利用NUTS参数优化方法实现了碳通量（Net Ecosystem Exchange,NEE）模型参数反演,并与Metropolis-Hastings（MH）方法进行对比。结果表明,本文算法的参数值达到稳定波动时的抽样次数减少了85%左右,参数优化效率提升3倍左右。参数优化后,2种NEE模型中7个参数不确定性降低10%~53%。此外,NEE模拟效果明显提升,模拟值与实测值的R²分别提高23%和17%,RMSE分别降低3%和4%。综上所述,本文提出的参数优化方法对生态领域的参数估计或数据同化工作具有一定的借鉴意义。