一种基于BP算法学习的小波神经网络

为发展 Szu的基于信号表示的小波神经网络 ,提出一种多输入多输出的小波网络模型 ,网络隐层采用框架小波函数、输出层采用 Sigmoid激励函数 ,并选用“熵误差函数”以加速网络的学习速度。奇偶判别和混沌时间序列预测例子的实验结果表明了它具有良好的函数逼近能力和推广能力 ,收敛速度和均方误差均优于相同结构的多层感知器模型。     

现场海域水平扩散系数的探讨 Ⅱ．黄河三角洲埕北海域的计算

根据Ｔａｙｌｏｒ的湍流扩散理论和Ｈａｙ、Ｐａｓｑｕｉｌｌ提出的拉格朗日流和欧拉流的相关函数的相似性，用黄河三角洲埕北海域的实测资料，按四种不同方案计算出该海域的水平扩散系数和谱密度分布。指出：该海区潮流扩散系数主要受潮流的影响，其分布特征与潮流的分布基本一致，所求的水平扩散系数主要是由周期性的潮运动给予潮流扰动所致。方案４是通过富氏变换滤波消去了周期性潮流扰动，本方案计算的水平扩散系数是由周期为６ｈ以下的小尺度涡提供能量。     

Generation of a double Kelvin wave resulting from the reflection of the Rossby wave by the jump layer

This paper considers the evolution of a spatially-localized divergent Rossby wave field near the depth jump. If the jump magnitude is comparable to the depth, Rossby waves are fully reflected and a double Kelvin wave is then generated. The Rossby waves and the double Kelvin wave are described by the first- and zero-approximation fields of the asymptotic expansion, respectively. Over the characteristic Rossby wave period, the level elevation produced by the double Kelvin wave spreads over an extensive area, theraby making up for the change in the total fluid mass of the Rossby waves.Translated by Vladimir A. Puchkin.     

Vertical habitat partitioning by large calanoid copepods in the oceanic subarctic Pacific during Spring

The copepods Neocalanus plumchrus, N. flemingeri, N. cristatus, and Eucalanus bungii dominate the net zooplankton throughout the subarctic Pacific Ocean. All four species have an extensive seasonal ontogenetic vertical migration, completing most or all of their feeding and somatic growth in spring and early summer. We used stratified tows with MOCNESS and BIONESS instrumented net systems to resolve their upper ocean vertical distributions in May and June of 1984, 1987 and 1988. In each year the feeding copepodite stages of all four species were concentrated above the permanent halocline (roughly from 0 to 150m). However, the four species showed strong vertical species zonation and segregation within this layer. We consistently found a near-surface pair (N. plumchrus and N. flemingeri) and a subsurface pair (N. cristatus and E. bungii). The boundary between these groups shifts vertically, but was sharply defined and was very often coincident with a weak and transient thermocline marking the base of the layer actively mixed by surface wind and wave energy. Diel vertical migration was very limited during our sampling periods.The data suggest that the vertical distribution patterns of the copepods could be set by responses to the local intensity of turbulent mixing in the watercolumn. N. plumchrus and N. flemingeri occupied a stratum characterized by strong turbulence. N. cristatus and E. bungii occupied a stratum that was a local minimum in turbulence profiles. The depth of the boundary between the species pairs was deeper when winds and surface energy inputs were strong. The vertical partition pattern may also be determined by a difference in feeding strategy between the species pairs. N. plumchrus and N. flemingeri may feed on the enhanced protozoan population of the mixed layer, while N. cristatus and E. bungii feed on particle aggregates settling from above.     

The impact of mesoscale eddies on plankton dynamics in the upper ocean

A coupled QuasiGeostrophic mixed-layer ECOsystem model (QGECO) is used to investigate the impact of the underlying mesoscale eddy field on the spatial and temporal scales of biological production and on overall rates of primary productivity. The model exhibits temporal trends in the biological and physical fields similar to those observed in the North Atlantic; i.e. the mixed layer shallows in spring causing a rapid increase in phytoplankton concentrations and a corresponding decline in nutrient levels. Heterogeneity is produced in the mixed layer through Ekman pumping velocities resulting from the interaction of windstress and surface currents. This variability impacts on biological production in two ways. Firstly, spatial variations in the depth of the mixed layer affect the photosynthetically active radiation (PAR) availability and hence production rates, and secondly, eddy enhanced exchange between the surface water and those at depth bring additional nutrients into the euphotic zone. These processes result in significant spatial and temporal heterogeneity in the ecosystem distributions.Investigation of the spatial heterogeneity of the biological system finds variability to be significantly greater than that of the mixed layer. The relationship between the eddy field and the ecosystem is investigated. The structure and correlation of the biogeochernical fields change with time. The biological fields are found to have a shorter horizontal scale, but whiter spectrum than the underlying eddy field.Overwinter conditions are found to have a profound effect on the variability, size and timing of the following spring bloom event. Variations in the nitrate levels are primarily responsible for the variability in the biological system in the first year. In subsequent years the variation in the overwintering population is found to be dominant.     

Interstitial water chemistry of the inner continental shelf sediments off the gironde estuary

The interstitial water composition ( , alkalinity, Ca²⁺, Mg²⁺, Sr²⁺, Na⁺, K⁺) and the cation exchange capacity (CEC) were determined for the muddy sediments of the continental shelf off the Gironde Estuary (France), in the area where the sediment represents the deposit of the muddy suspension of the river. In comparison with seawater concentrations, the pore waters below 10 cm depth, show depletions of and Ca²⁺ and below a 30 cm depth show depletions of Mg²⁺. Inversely, the upper 10 cm an enrichment of Ca²⁺ concentration, and an increase of K⁺ concentration to a 40 cm depth. High values of are observed at the top 4 cm. Alkalinity enrichment is observed along the length of the core. Applying the alkalinity models for the sediment below a 10 cm depth demonstrates generally that calculated alkalinities are higher than the measured ones. Ca²⁺ dissolution occurs at the first 10 cm and authigenic carbonate precipitation starts beneath that level. Mg²⁺ depletion is accompanied by bicarbonate loss. This proves that Mg²⁺ depletion is due to a Mg-silicate reaction. The result of the CEC does not confirm the Mg²⁺ uptake by clay minerals in exchangeable site, under reducing conditions. Diffusion and bioturbation play an important role in the pore water concentration at the top of the core.