基于混合像元分解和知识规则的人工刺槐林健康分类

提出一种基于精度较高的Landsat TM数据检测刺槐林枯梢状况的方法,整个过程分为:1)利用监督分类提取出林地分布范围,制作刺槐林地掩膜;2)对刺槐林地利用线性光谱混合模型进行混合像元分解,得到健康刺槐、非光合作用绿色植被(NPV)、土壤的丰度图,定义刺槐林的健康分类等级:健康或轻度枯梢、中度枯梢、重度枯梢或死亡;3)利用决策树分类器对丰度图像、归一化植被指数(NDVI)以及经缨帽变换获取的亮度和绿度成分定义一组规则,进行刺槐林生长健康状况分类,分类结果与实地样方数据具有较好的一致性,利用林冠分级向导进行精度分析,精度达82.2%,Kappa系数为0.752。     

Water discharge variability of Changjiang （Yangtze） and Huanghe （Yellow） Rivers and its response to climatic changes

Influences of large-scale climatic phenomena, such as the E1Nifio/La Nifia-Southem Oscillation （ENSO） and the Pacific Decadal Oscillation （PDO）, on the temporal variations of the annual water discharge at the Lijin station in the Huanghe （Yellow） River and at the Datong station in the Changjiang （Yangtze） River were examined. Using the empirical mode decomposition-maximum entropy spectral analysis （EMD- MESA） method, the 2- to 3-year, 8- to 14-year, and 23-year cyclical variations of the annual water discharge at the two stations were discovered. Based on the analysis results, the hydrological time series on the inter- annual to interdecadal scales were constructed. The results indicate that from 1950 to 2011, a significant downward trend occurred in the natural annual water discharge in Huanghe River. However, the changes in water discharge in Changjiang River basin exhibited a slightly upward trend. It indicated that the changes in the river discharge in the Huanghe basin were driven primarily by precipitation. Other factors, such as the precipitation over the Changjiang River tributaries, ice melt and evaporation contributed much more to the increase in the Changjiang River basin. Especially, the impacts of the inter-annual and inter-decadal climate oscillations such as ENSO and PDO could change the long-term patterns of precipitation over the basins of the two major rivers. Generally, low amounts of basin-wide precipitation on interannual to interdecadal scales over the two rivers corresponded to most of the warm ENSO events and the warm phases of the PDO, and vice versa. The positive phases of the PDO and ENSO could lead to reduced precipitation and consequently affect the long-term scale water discharges at the two rivers.     

Lopingian mixed floras from Linxi Formation in Soron area,Inner Mongolia

A lot of well-preserved plants from the Linxi Formation are collected in the Soron area,Inner Mongolia,consisting of 34 species of 16 genera. They are Late Permian in age. The flora is characterized by a lot of Angaran plants,such as Paracalamites,Comia,Rhachiphyllum,Iniopteris,Rufloria,etc. Some Cathaysian elements,such as Lobatannularia lingulata,L. multifolia,Fascipteris Cathaysiantus,F. densata,Taeniopteris szei,Cladophlebis liulinensis and C. permica,are also mixed within the present flora. This indicates that the Soron of Inner Mongolia was located in the sector between the Angaran and Cathysian floristic provinces in Lopingian. It is beneficial for depicting the location and the evolution of the Solenker--Xar Moron suture zone in the phytogeographic view. The transmigration of tropical and subtropical Cathaysian plants to the north indicates that the Paleoasian Ocean was closed during the late Permian along the suture zone.     

复杂机场环境的多维动态可视化方法

针对飞行器可视导航中复杂机场空地环境多维要素动态交互可视化的难题,基于空地环境要素威胁语义统一的信息场模型,采用体绘制、面绘制与切片绘制混合的方法对机场近区域复杂多样的地形地貌、电磁和气象等威胁态势信息进行交互式动态可视化表达。以某机场的典型模拟数据为例,实现了威胁态势的动态交互可视化,验证了该方法的实用性和有效性。     

基于ROMS模型数值研究南海温跃层的季节变化

On the basis of the regional ocean modeling system （ROMS）, the seasonal variations of the thermocline in the South China Sea （SCS） were numerically investigated. The simulated hydrodynamics are in accordance with previous studies： the circulation pattern in the SCS is cyclonic in winter and anticyclonic in summer, and such a change is mostly driven by the monsoon winds. The errors between the modeled temperature profiles and the observations obtained by cruises are quite small in the upper layers of the ocean, indicating that the ocean status is reasonably simulated. On the basis of the shapes of the vertical temperature profiles, five thermocline types （shallow thermocline, deep thermocline, hybrid thermocline, double thermocline, and multiple thermocline） are defined herein. In winter, when the northeasterly monsoon prevails, most shallow shelf seas in the northwest of the SCS are well mixed, and there is no obvious thermocline. The deep region generally has a deep thermocline, and the hybrid or double thermocline often occurs in the areas near the cold eddy in the south of the SCS. In summer, when the southwesterly monsoon prevails, the shelf sea area with a shallow thermocline greatly expands. The distribution of different thermocline types shows a relationship with ocean bathymetry： from shallow to deep waters, the thermocline types generally change from shallow or hybrid to deep thermocline, and the double or multiple thermocline usually occurs in the steep regions. The seasonal variations of the three major thermocline characteristics （the upper bound depth, thickness, and intensity） are also discussed. Since the SCS is also an area where tropical cyclones frequently occur, the response of thermocline to a typhoon process in a short time scale is also analyzed.     

短期风化对混合溢油组成的影响

对两种原油进行混合配比实验,将混合后油样置于人工气候箱进行风化模拟实验,采用气相色谱质谱联用仪(GC-MS)检测风化样品中的生物标志化合物,短期风化作用对混合溢油的油指纹、生物标志化合物诊断指标等的影响。结果表明,混合溢油的正构烷烃总质量变化与单一原油油品的变化规律相近,即前期风化较快,质量减损较多,而后期风化趋缓,不同混合比例的溢油表现差异不明显。常用于短期风化的诊断比值对各混合油样的风化具有指示意义,但难以定性鉴别油品是否发生混合。重复性限法检验只能判定短期风化过程前后的油样为同一油源,而难以反映油样是否为混合油源的特征,各诊断比值的RSD%值较大可能是混合溢油的一个表现。     

区域Gamma混合模型的SAR图像分割

针对传统Gamma混合模型用于SAR图像分割时忽略像素间空间相关性,导致分割结果不连续并产生大量误分割的现象,提出了区域Gamma混合模型的SAR图像分割算法。首先对图像进行分水岭分割,得到过分割区域块,然后将其作为输入样本进行基于Gamma混合模型的聚类,在模型的参数估计过程中进一步考虑区域间的空间相关性,设计邻域因子融入到迭代过程,得到邻域加权类分布概率。该算法充分利用像素间的空间相关性,能够降低噪声对分割结果的影响。通过合成图像和真实SAR图像的实验表明,本文算法能够实现SAR图像的准确分割。     

The roles of different mechanisms related to the tide-induced fronts in the Yellow Sea in summer

In summer, the Yellow Sea Cold Water Mass （YSCWM） is a stable water mass of low temperature lying at the bottom of the central Yellow Sea （YS）. It is fringed by some typical tidal fronts, which separate deep, stratified water on the offshore side from the well-mixed, shallow water on the inshore side. Three striking fronts--Subei Bank Front （SBF）, Shandong Peninsula Front （SPF）, and Mokpo Front （MKF; a front off the southwestern tip of the Korean Peninsula）--have been iden- tified by various studies from both satellite observations and model results. Tide plays an important role in the formation and maintenance of these fronts. However, it is still a matter of debate as to the roles these two kinds of mechanisms of upwelling and tidal mixing play, and how importance they are in the maintenance processes of the above three fronts. Basing a nested high-resolution model HYCOM （the Hybrid Coordinate Ocean Model）, this study focuses on the different mechanisms of tidal effects on the thermal fronts in the YS in summertime. Through comparative experiments with and without tidal forcing, the results indicate that the MKF is mainly driven by tide-induced upwelling. For the SPF, tidal mixing is the dominant factor, when lower cold water is stirred upwards along the sloping topography of the western YS. Meanwhile, the combined effect of upwelling and tidal mixing is the main cause of the formation of the SBF. Diagnostic analysis of thermal balance shows that horizontal nonlinear advection induced by strong tidal currents also contributes to the thermal balance of frontal areas.     

被动微波遥感雪深反演混合像元问题研究EI北大核心CSCD

积雪深度是反映地表积雪量变化的重要因子,是水文模型和气候模式中的重要参数之一。被动微波遥感以其穿云透雾、对雪层信息敏感等特点,被广泛应用于雪深的反演研究中。被动微波传感器的低空间分辨率(数千米至数十千米)及地表覆盖的复杂性,使星载被动微波遥感影像中的混合像元现象十分突出,严重制约了被动微波雪深遥感监测的应用和发展。目前,被动微波遥感雪深反演中的混合像元问题研究仍存在着诸多挑战:①理论支撑不足,对被动微波混合像元亮温响应特征及影响机制的研究尚不充分;②针对混合像元问题所发展的被动微波雪深反演算法,对地表异质性特征考虑不足。     

加乘性混合误差模型精度评定的SUT法EI北大核心CSCD

已有加乘性混合误差模型参数估计方法能达到二阶精度,但精度评定方法只能达到一阶精度,若通过传统泰勒级数展开近似函数法来获取参数估值的二阶精度信息,由于加乘性混合误差模型中参数估值与观测值为一个复杂的非线性关系,必然需要复杂的求导运算。针对该问题,本文使用一种无须求导、无须了解非线性函数构成的比例无迹变换(scaled unscented transformation,SUT)法来计算参数估值的二阶精度信息。通过算例分析表明,利用SUT法求解加乘性混合误差模型能够有效避免复杂的求导运算,所求得的参数估值及其协方差阵均能达到二阶精度,从而验证了本文方法的可行性和优势。