Regionalization of unit hydrograph Parameters: 2. Uncertainty analysis

Hydrologic model parameters obtained from regional regression equations are subject to uncertainty. Consequently, hydrologic model outputs based on the stochastic parameters are random. This paper presents a systematic analysis of uncertainty associated with the two parameters, N and K, in Nash's IUH model from different regional regression equations. The uncertainty features associated with N and K are further incorporated to assess the uncertainty of the resulting IUH. Numerical results indicate that uncertainty of N and K from the regional regression equations are too significant to be ignored.     

Barotropic local instability and severe storm process

By means of barotropic model, the characteristic and initial value problems are investigated to reveal the local two-dimensional barotropic instability of the nonuniform current to the dynamic mechanism of the formation of the Yangtze-Huaihe River severe storm in July 1991. Analytical theory and numerical experiment show that (i) the unstable developing modes are chiefly the two periods of about 44 d and 10 d, which are fundamentally consistent with that of the precipitation change of the Yangtze-Huaihe River. (ii) The growth rate of the local perturbation is dominated by the meridional wave numbern = 1–5 and zonal wave numberk = 1–12, i.e. the severe storm over the Yangtze-Huaihe River results from the interaction of the systems at different latitudes and waves of different scales, (iii) The perturbation over the Yangtze-Huaihe River possesses the property of local intensification, which slowly migrates westward over the lower and middle reaches of the Yangtze-Huaihe River. (iv) The growth rate of the instability, especially the propagation velocity of the perturbation, is sensitive to the external parameters ū and α. Project supported by the National Natural Science Foundation of China.     

Properties of wave velocity for two types of granitoids at high pressure and temperature and their geological meaning

The wave velocity for two types of granitoids was measured using the analytic method of full-wave vibration at high pressure and high temperature. The laws of velocity changes for them differ with the pressure boost and temperature rise, and the velocity change of S-type is more violent than that of I-type. The “softening point” of compressional wave velocity (V μ) is also revealed during the measurement for two types of granitoids imitating the pressure and temperature at a certain depth. But the depth of “softening”, V_p after “softening” and the percentage of V_p’s drop around the “sofrening point” for two types of granitoids are obviously different. The depth of “softening” is 15 km approximately and V_p after “softening” is 5.62 km/s for S-type granitoid. But for I-type granitoid the depth of “softening” is 26 km approximately and V_p after “softening” is 6. 08 km/s. Through careful analysis of rock slices after the experiment, it was found that the “softening” of elastic-wave velocity is caused by the partial melting of granite. Combined with the results of geophysical prospecting, these results suggest that the low-velocity layers developing in the interior of Earth crust are related to thc partial melting of different types of granitoids. The formation of the low-velocity layer in the upper-middle Earth crust is closely related to the development of S-type granitoid, but that in the lower Earth crust is closely related to the development of I-type granitoid.     

Regionalization of unit hydrograph Parameters: 2. Uncertainty analysis

Hydrologic model parameters obtained from regional regression equations are subject to uncertainty. Consequently, hydrologic model outputs based on the stochastic parameters are random. This paper presents a systematic analysis of uncertainty associated with the two parameters, N and K, in Nash's IUH model from different regional regression equations. The uncertainty features associated with N and K are further incorporated to assess the uncertainty of the resulting IUH. Numerical results indicate that uncertainty of N and K from the regional regression equations are too significant to be ignored.     

图像配准和融合及其在医学影像中的应用

图像配准和融合是图像分析和处理的基本问题，在医学影像、遥感、计算机视觉等领域有着广泛的应用。本文研究它们在多模态医学影像技术中的应用。本文提出并实现了基于Ｆｏｕｒｉｅｒ变换的图像配准和基于子波变换的图像融合方法，在图像融合中，我们采用特征选择的取大原则，这种准则更适合于来自不同图像中的明显特征     

黄河上游径流预报的灰色拓扑方法

以龙羊峡水库年平均入库流量预报为例，根据灰色系统理论的建模方法，利用龙羊峡水库入库水量代表站唐乃亥水文站的实测径流资料，建立了一个ＧＭ（１，１）拓扑预测模型，用于黄河上游径流的长期预报，并取得了较为满意的结果。     

区域采样法测定水中硫酸根离子的研究

在分散度不同的区域内进行区域采样用于水中硫酸根离子的测定，结果令人满意。应用此方法测定硫酸根离子的线性范围为２００．０～３０００ｍｇ／Ｌ；进样频率为１２０次／ｈ；相对标准偏差为１．７０％     

关于大陆与陆缘的深部地震探查的进展──1994、1996年两届国际学术讨论会综述

连续两届国际深部地震讨论会成果累累。对深部地震的资料采集、处理、解释技术以及几类专门地质地球物理问题的研究进行了综述。研究成果在理论、方法方面是该领域的世界先进水平：在地震学应用于陆缘、陆壳与壳幔过渡带等地质问题方面有明显的提高。提出了值得讨论的问题，最后探讨了发展趋势     

地质样品中微量金银的在线分离富集方法

提出用ＶＳ－Ⅱ型强碱性阴离子交换纤维定量富集、硫脲解脱，流动注射在线分离富集—火焰原子吸收光谱法测定地质样品中痕量金银的分析方法。该方法检出限低，金为１．４μｇ／ｍＬ，银为０．３５μｇ／ｍＬ，相对标准偏差均小于２％，采样频率为１２０次／ｈ，用于地质样品中痕量金银的直接测定，结果令人满意。     

苏州地区I-A型花岗岩特征与成因的对比研究

应用岩石学、矿物学、地球化学及岩浆物理学等的理论和方法，对苏州地区Ｉ型和Ａ型花岗岩的特征及成因进行了对比研究，得出以下主要结论：Ｉ型花岗岩为晚侏罗世火山弧压性环境产物，侵位较深；而Ａ型者为Ａ２型花岗岩，为早白垩世造山后张性环境产物，侵位较浅，具晶洞构造．Ａ型比Ｉ型花岗岩以酸度、碱度、Ａｌ，Ｆ含量高而Ｈ２Ｏ＋含量低，以ＲＥＥ及不相容元素丰度高，负Ｅｕ异常强，而ｗ（ＬＲＥＥ）／ｗ（ＨＲＥＥ）低为特色．Ａ型者矿物结晶顺序先浅色后暗色，暗色矿物以富铁贫镁、ＲＥＥ分配系数高及负Ｅｕ异常明显及长石富碱贫钙，而与Ｉ型花岗岩不同．Ａ型者副矿物４０余种，属锆石－褐帘石－萤石型，锆石结晶温度高，（１１０），（１１１）晶面发育；而Ｉ型者仅１０余种，属榍石－锆石－磷灰石型，锆石结晶温度低，（１００），（３１１）晶面发育．Ｉ型花岗岩为幔源岩浆与下地壳部分熔融岩浆混合及ＡＦＣ作用产物，而Ａ型花岗岩为残余下地壳（麻粒岩相）熔融后分离结晶产物，二者并非同源岩浆演化关系．Ａ型比Ｉ型岩浆来源浅，酸度大，含水少；故岩浆密度小，粘度大，上升慢，冷凝长．受密度筛影响，Ａ型花岗岩岩浆房及侵位深度小于Ｉ型