基于组合核非线性退化模型的遥感图像复合分类

遥感数据的多空间分辨率复合分析是遥感处理技术的重要发展方向。为了解决低分辨率图像混合像元分类精度低、高分辨率数据分类处理时间长以及大区域高分辨率数据获取困难等实际应用问题,该文改进了传统基于线性退化函数模型的复合分类模型,提出了基于组合核函数的非线性退化模型复合分类算法,分析了纹理信息对于提高复合分类精度的作用,并通过实际遥感数据试验分析比较了两种模型的分类精度。试验结果表明新方法可较大程度地提高总体分类精度,在分类过程中引入纹理信息有助于进一步改善分类精度。     

不同生长期小麦茎秆力学性质与形态特性的相关性（简报）

为了研究小麦茎秆生物力学性质指标与小麦优种特性的相关性,以正在优种选育过程中的4个北方小麦品种为研究对象,试验测定了开花期、乳熟期及蜡熟期茎秆基部第二节间的惯性矩、拉伸强度、弯曲强度、剪切强度、弹性模量、抗弯刚度等生物力学指标及小麦形态、生理特性指标值;应用SAS软件进行相关性分析,筛选出了开花期主要生物力学评价指标为弯曲强度,乳熟期为抗弯刚度.并且确定了与相应生物力学相关度较大的形态、生理特性指标,开花期为:外径、茎秆干质量、株高、节间距、含水率、壁厚、穗质量;乳熟期为穗质量、外径、株高、叶鞘高、茎秆干质量.并且随着生长期的不同,相关度也不同,尤其是与含水率的相关度逐渐减小.该研究为建立生物力学指标评价体系提供了选择依据.     

不同条件下夏玉米冠层反射光谱响应特性的研究

通过对两个年份不同牛育阶段的田间夏玉米活体进行冠层反射光谱测定,该文分析了不同空间尺度、种植密度、观测角度、生育阶段、叶面积指数KAI、氮素胁迫、叶片含水量以及与杂草共生等8个条件下的冠层反射光谱响应特性,并探讨其发生机理,研究结果表明,夏玉米在不同条件下的冠层反射光谱响应均呈现出一定的规律性.在近红外波段,其反射率值随氮肥施用量的增加而增大,尤以750～1350 nm波段,但在可见光波段,反射率值降低;随着观测探头的逐步升高,反射率值降低:随杂草量的增多,反射率值逐渐增大.这些结果揭示了高光谱遥感田间夏玉米理化信息获取的巨大潜力,同时,为以后遥感反演建模提供了依据.     

有机修饰塿土表面特性的研究Ⅰ.CEC和比表面

污染物在土壤中的行为与土壤的表面化学特性密切相关,研究土壤的表面化学特性对于从微观领域了解污染物行为及其变化规律,采取有效措施减轻并消除土壤中污染物的危害,保护土壤环境和人体健康具有重要意义。为了增强土壤对有机污染物的吸附固定,国内外开始研究应用阳离子型表面活性剂对黏土矿物或土壤进行修饰,使土壤的表面性质由亲水性转变为疏水性,增大对土壤中有机污染物的吸附能力[1,2],研究表明,利用阳离子型表面活性剂修饰黏土矿物或土壤,可以显著增强土壤对水中有机污染物的吸附固定能力[3~7],但对于重金属离子却具有不同的作用[8~11]。土壤胶体的比表面、离子交换能力是与土壤对污染物吸持直接相关的土壤表面性质     

Adsorption,Desorption And Mobility of Fomesafen in Chinese Soils

Fomesafen, 5-[2-chloro-4-(trifluoromethyl)phenoxy]-N-(methylsulfonyl)-2-nitrobenzamide, is used widely for weed control in soybeans since its introduction to China. Little information is available on its adsorption, desorption and movement in Chinese soils. The adsorption, desorption and mobility of fomesafen in six Chinese soils was studied. Adsorption isotherms agreed with the Freundlich equation very well. The results of regressionanalysis indicated that soil pH was more important than organic matter for fomesafen adsorption. Fomesafen was more readily desorbed from soils with 0.01 M CaSO₄ solution. Soil TLC and column leaching studies showed that fomesafen and its metabolites was less mobile in Chinese soils. About 89.82% of applied fomesafen and its metabolites still remained in upper 5 cm layer 60 days after treatment under field conditions.     

Spatial pattern and heterogeneity of soil properties in sand dunes under grazing and restoration in Horqin Sandy Land, Northern China

Applying a combination of classical and geostatistical methods, we identified soil properties and their spatial variation in a 5-year grazed sand dune (GSD5) and a 20-year recovered sand dune (RSD20) in Horqin Sandy Land, northern China. The paper assesses the effect of grazing, topography and vegetation restoration on spatial heterogeneity of soil properties. The results showed that soil organic carbon, total nitrogen, very fine sand (0.1–0.05 mm) content and their coefficients of variation were lower in GSD5 than in RSD20, while soil water contents (0–20 cm and 20–40 cm depths) were higher in GSD5 than in RSD20. Geostatistical analysis revealed that the spatial structured variance accounted for the largest proportion of total sample variance in soil properties at the measured scale under grazing and restoration. The spatial autocorrelation ranges were 66.30 m for soil organic carbon and 50.80 m for total nitrogen in GSD5 less than those in RSD20 (70.00 m and 76.10 m, respectively), while the spatial autocorrelation ranges of soil particle size fractions and soil water contents in RSD20 were less than those in GSD5. Kriging-interpolated maps also showed that the heterogeneity of soil organic carbon and total nitrogen and their degree of patch fragmentation were higher in GSD5 than in RSD20. These results suggested that continuous grazing resulted in an increase in spatial variability of soil nutrient and a decrease in spatial variability of soil particle size fractions and soil water content. Soil organic carbon and total nitrogen of sand dunes are associated closely with soil particle size fractions, relative height of sampling site and vegetation cover. Spatial patterns of soil properties are most strongly related to grazing, topography and plant-induced heterogeneity in sand dune ecosystems prone to wind erosion.     

流动沙丘风沙流结构的定量研究

在对野外风沙流进行实测的基础上,运用拟合曲线、拟合方程等方法对0～48cm垂直高度层内风沙流结构进行了分析研究。结果表明：0～48cm高度内总输沙率及各高度层输沙率均与风速呈幂函数关系;各高度层输沙率百分比与风速呈指数函数关系;各高度层输沙率与高度呈幂函数或指数函数关系;上、中、下各层输沙率与总输沙率呈良好的幂函数关系。     

基于DEM的马莲河流域数字地形分析

研究基于DEM的数字地形分析可以为各个领域提供地形分析的基础理论及实现手段。DEM水平分辨率作为重要的尺度参数,是确定地形参数和应用尺度的重要指标。选取陇东黄土高原地区的马莲河流域作为研究对象,应用美国ESRI公司开发的ArcGIS地理信息系统软件,基于数字高程模型进行流域地形特征的提取与分析。研究提取流域的基本地形特征如坡度和坡向,提取流域水系,划分了小流域。通过分析比较,得出DEM水平分辨率对提取的地形特征及水文特征的影响,为马莲河流域的数字地形分析提供了理论依据。     

Genetic Relationships of some PCNA Persimmons (Diospyros kaki Thunb.) from China and Japan revealed by SRAP analysis

Sequence-related amplified polymorphism (SRAP) marker was employed to analyze genetic relationships of 10 Chinese, 11 Japanese varieties of persimmon and its 6 relatives. Twenty out of 72 primer combinations could amplify clearly and consistently. Just only one primer set could discriminate all the genotypes used in this study. The genetic relationships were analyzed using unweighted pair-group method of arithmetic average cluster analysis (UPGMA) and principal coordinated analysis (PCOA). The results showed: Chinese varieties were grouped into two clusters, while the Japanese varieties were grouped into one cluster with two subgroups. The persimmon relatives were clustered together. Probably, Chinese and Japanese Pollination-constant and Non-astringent (PCNA) persimmons had different genetic background. ‘Eshi No.1’ from Hubei province was close to Male type 6. ‘Xiangxitianshi’ from Xiangxi municipality, Hunan province, had a high similarity coefficient with ‘Maekawa-Jirou’ and was clustered in the Japanese varieties group, indicating that ‘Xiangxitianshi’ maybe close to the Japanese variety in terms of genetic background. The Chinese Pollination-constant and Astringent (PCA) and PCNA persimmons were not distinctly distinguished implying that the non-astringent trait of Chinese PCNA type might result from the PCA type.     

Characterization of the chemical composition of lotus plumule oil

Characterizations of lotus plumule and plumule oil, focusing on approximate composition analysis of lotus plumule powder and fatty acid composition, lipid classes, triglyceride (TG) profiles, and sterol analysis of the plumule oil, were conducted in this work. The results revealed that the lotus plumule constitutes 7.8% moisture, 4.2% ash, and 12.5% crude oil and 26.3% protein on the dry base. Lotus plumule oil is rich in linoleic acid (50.4%) and oleic acid (13.5%), and the dominating saturated fatty acids are palmitic acid (18.0%) and behenic acid (6.8%). The principal components of TG in lotus plumule oil are LLL (12.80%), beta-PLL (11.27%), beta-POL (8.28%), beta-PLO (8.58%), and beta-BeLL (8.32%). Lipid class assay of the crude oil gave the saponification value of 153.4 KOH mg/g and tocopherol content 390 mg/100 g. A distinct characteristic of lotus plumule oil is that its unsaponifiable matter is incredibly high, up to 14-19%, which consists mainly of beta-sitosterol (32%), Delta(5)-avenasterol (20%), and campesterol (6.3%). The major occurring form of sterols was found to be steryl ester. This work might be useful to develop innovative applications of lotus plumule oil.