南海月平均流的计算

本文用一个二维全流方程模式,计算了南海1—12月的逐月平均流。所得的南海逐月环流的总趋势以及一些中小尺度的涡旋现象不仅与海洋观测相一致,也与已有的一些研究结果相吻合。而且,计算所得的海洋场和海面起伏场均很好地反映出南海的季节性变化特征,特别是高海暖流的存在在计算中又进一步得到了证实。文中对南海海流的机理也作了一此初步探讨。     

矿物分类和识别的人工神经网络模型

本文提出矿物分类和识别的人工神经网络模型，并选取一组标样－我国沉积碳酸盐型锰矿中菱锰矿作为研究对象，识别效率达１００％。结果表明，该模型性能良好，可望成为矿物识别的有效手段。     

运用人工神经网络作汛期降水预报

本文提出汛期降水预报的人工神经网络方法，并选择一组标样进行了具体分析，预报成功率达100％，结果表明，该方法性能良好，可望成为汛期降水预报的有效手段。     

硅酸盐类样品中硅、铁、铝、钛、锰、钙、镁、钾、钠和磷的x射线荧光光谱测定

本文拟定了一种以熔融法制备样片,用X射线荧光光谱测定硅酸盐类样品中Si、Fe、Al、Tj、Mn、Ca、Mg、K、Na、P等元素的分析方法。在对不同靶材X光管和分光晶体实验对比的基础上,选择了最佳的测量条件。该法具有快速、准确,测量范围广,检测限低,价格便宜等优点。经过近百个各种类型标样或管理样品的分析对比表明,本法不仅适用于硅酸盐类岩石样品的分析,还适用于铁矿、铝土矿、碳酸盐类岩矿样品以及水泥、耐火材料等样品的分析。     

南岭山地湖泊多环芳烃的大气沉降历史记录

高大山脉对大气环流具有阻挡或屏障作用.通过分析测定南岭山地湖泊沉积物中多环芳烃(PAHs)的垂直分布,结合210Pb定年,探讨了大气PAHs污染在南岭山地的沉积历史.结果显示,沉积物中的PAHs主要以低环数化合物为主,其中尤以菲的含量为最高.在剖面深度0～28 cm范围内,总多环芳烃的含量范围为86～778 ng/g.自1970年开始,PAHs含量持续增加,其中20世纪80、90年代中后期PAHs含量略有降低;在2000年后,沉积柱中的PAHs含量呈急剧增加之势态.多环芳烃在大气迁移过程中发生了组成分异,沉积物中相对富集轻组分(低环数)PAHs,与颗粒物结合的大气干湿沉降是PAHs向偏远山地积的主要途径.     

黄土孔隙性分类判别的人工神经网络方法

运用人工神经网络的一典型模型——“反向传播”神经网络,对洛川黄土孔隙性的实测数据进行了分析,建立了洛川黄土孔隙性预测的计算机智能专家系统。结果表明,神经网络方法性能良好,可望成为黄土孔隙性分类、判别的有效辅助手段。     

Evaluation of aluminum speciation in surface waters in China and its environmental risk assessment

As the ongoing global research on acid precipitation is developing in depth, more and more attention has been paid to the ecological effects of aluminum (Al) due to its toxicity to plants and animals, which is caused by acid precipitation. As a very serious problem of terrestrial and aquatic environmental acidification occurs in China, especially in southwestern China, a systematic investigation of Al speciation in these regions is very important. In this paper, the Al speciation results of surface waters in China are reported and its ecological impacts is evaluated. More than 100 water samples were collected from about twenty provinces of China. Driscoll's Al speciation scheme combined with the modified MINQEL computer model is used for speciation of Al. This study shows that the ecological impacts of acidification are quite different between China and Western countries, because of different geographical environments and geological settings. In Western countries, acidification is mainly caused by NO₂^-. Due to low concentrations of K⁺, Na⁺, Ca²⁺, Mg²⁺, the buffer capacities of soil and water are weak. Therefore, natural waters can be acidified to pH<5 very easily, resulting in a considerable mobilization of Al and worsening of the ecological environment. In China, acid precipitation is mainly in the form of sulfuric acid. In northwestern China, concentrations of K⁺, Na⁺, Ca²⁺, Mg²⁺ are high in soil and surface waters. This leads to much higher capacity and a high resistance ability to acidification. The pH values of waters in this region are high (around 7) and no serious Al toxicity is found at present. However, in northeastern and southeastern China, the soil is rich in Al (unsaturated aluminosilicates in northeastern China, saturated aluminosilicates in north and central China, aluminum-rich soil in southeastern and southwestern China). The concentrations of K⁺, Na⁺, Ca²⁺, Mg²⁺ in soil and waters are lower than those of northwestern China. Therefore the buffer capacity is limited. Numerous surface waters have already been acidified and pH values declined to 5. The impacts of Al toxicity on ecological systems in these regions are very serious, especially in Jiangxi, Hubei Provinces and Chongqing Municipality.     

Spatial and temporal trend of sewage pollution indicated by coprostanol in Macao Estuary,southern China

Bottom sediments and a ²¹⁰Pb dated sedimentary core in Macao Estuary were quantitatively analyzed with GC-MS for coprostanol and other sterols. Higher coprostanol concentrations were detected at Port Interior and Lower Qianshan River, indicative of serious sewage pollution from densely populated Macao Island and Zhuhai City. However, very low concentrations of coprostanol were detected around Coloane Island as well as Port Exterior and Maliuzhou River except sites adjacent sewage outlets. Coprostanol pollution in Macao Estuary originates mainly from locally direct discharge of untreated wastewater. The concentrations of coprostanol in ZJ-9 were mainly in range of 150–280 ng/g with an average of 210 ng/g for more than 20 years from early 1970s to early 1990s. However, it increased obviously since 1993, and reached the highest in 1995–1996. However, a sharp decrease of coprostanol concentration from 470 to 31 ng/g after 1996 was observed corresponding to the first wastewater treatment plant in Taipa Island going into action.     

ON BROADENING MECHANISMS OF THE Ca XIX RESONANCE LINE IN SOLAR FLARES

There are two explanations for the broadening mechanisms of the Caxix resonance line in the rising phase of a solar flare. One is due to the velocity dispersion of the chromospheric evaporation along the loop; the other is due to the non-thermal turbulence resulting from the flare energy release. In this paper we distinguish between these two possibilities by studying the influence of the loop orientation on the broadening of the Caxix resonance line, based on the data observed with SXT and BCS on Yohkoh. The results seem to support that the broadening of the Caxix resonance line is caused by non-thermal turbulence, rather than by the velocity dispersion.     

A modified ISBA surface scheme for modeling the hydrology of Athabasca River Basin with GCM-scale data

A soil–vegetation–atmosphere transfer model (SVAT), interactions between the soil–biosphere–atmosphere (ISBA) of Météo France, is modified and applied to the Athabasca River Basin (ARB) to model its water and energy fluxes. Two meteorological datasets are used: the archived forecasts from the Meteorological Survey of Canada’s Global Environmental Multiscale Model (GEM) and the European Centre for Mid-range Weather Forecasts global re-analysis (ERA-40), representing spatial scales typical of a weather forecasting model and a global circulation model (GCM), respectively. The original treatment of soil moisture and rainfall in ISBA (OISBA) is modified to statistically account for sub-grid heterogeneity of soil moisture and rainfall to produce new, highly non-linear formulations for surface and sub-surface runoff (MISBA). These new formulations can be readily applied to most existing SVATs. Stand alone mode simulations using the GEM data demonstrate that MISBA significantly improves streamflow predictions despite requiring two fewer parameters than OISBA. Simulations using the ERA-40 data show that it is possible to reproduce the annual variation in monthly, mean annual, and annual minimum flows at GCM scales without using downscaling techniques. Finally, simulations using a simple downscaling scheme show that the better performance of higher resolution datasets can be primarily attributed to improved representation of local variation of land cover, topography, and climate.