祁连山中段青海云杉林土壤肥力质量评价研究

以祁连山排露沟流域青海云杉林为研究对象,研究了海拔梯度上土壤肥力因子的分布特征及变化规律,并运用主成分分析法对青海云杉林土壤肥力状况进行了评价。结果表明：（1）研究区土壤呈碱性,pH值均大于8.0;高海拔地区（3 300 m）含水量达到过饱和状态,各土层含水量均大于100%;随海拔升高,全氮含量呈增大趋势,全钾含量呈减小趋势,而全磷含量呈先减小后增大趋势;不同海拔梯度速效磷含量差异不显著（P>0.05）,海拔3 300 m处速效钾含量显著高于其他海拔段（P<0.05）。（2） 不同海拔梯度下土壤有机质、全氮、全磷、全钾、速效磷和速效钾含量都有明显的“表聚效应”,其中3 300 m处0~10 cm土层有机质含量高达325.93 g·kg^-1,是本海拔段其他土层的1.6~1.8倍,是同土层其他海拔段的1.3~2.0倍。（3） 土壤肥力因子间关系密切,土壤含水量与有机质、全氮呈极显著正相关关系,与土壤容重、pH和全钾呈极显著负相关关系,土壤养分含量之间存在不同程度的显著正相关关系。（4） 不同海拔梯度土壤肥力质量为：3 300 m>3 200 m>3 100 m>3 000 m>2 900 m。     

中国市域居民住房支付能力时空演变及影响因素

利用泰尔指数、σ系数、莫兰指数和空间回归模型等方法,对2012—2017年中国大陆346个市域单元收入房价比的时空演变规律进行分析,并探究其影响因素。结果表明:2012—2017年,中国市域单元收入房价比排名整体表现为东部>中部>东北>西部,其相对差异和绝对差异均随时间呈现增大趋势,居民住房支付能力稳步增强。2012—2017年,中国市域单元收入房价比的空间集聚性显著增强,空间分布不均衡性有逐渐增强的趋势。经济水平、城镇化水平、房地产投资水平、城市规模是影响中国市域住房支付能力的主要影响因素,其中,经济水平和城镇化水平对居民住房支付能力具有较大的正向影响,而房地产投资水平和城市规模为负向影响;2012—2017年我国东北和东中西部则不同程度上受不同影响因素的影响。     

城镇增长下的收缩：以武汉为例

首先从人口、经济、用地3个维度综合考察武汉市增长与收缩的全貌,并采用县区及街道2个尺度的数据定量描述了武汉市增长与收缩的特征与空间格局,发现武汉市下辖青山区、硚口区、汉阳区和蔡甸区存在局部较严重的收缩现象,空间上形成集聚,形态上呈“穿孔式”。进一步以青山区为案例,着重从资本视角探讨发生局部收缩的内在机制,发现其存在老龄化、少子化趋势,但局部收缩的主因是资本从产业部门的“逃逸”。     

高分六号遥感卫星新增波段下的树种分类精度分析

高分六号卫星具有覆盖广、多种分辨率、波段多的优势,能为遥感解译提供更丰富的信息。为探究高分六号卫星新增波段在森林树种识别上的应用,本文以覆盖根河市阿龙山林业局的一期高分六号宽幅影像为数据源,基于特征优化空间算法（Feature Space Optimization,FSO）和最大似然分类法,分别利用高分六号的前4个波段和所有波段（8波段）的光谱、纹理等特征进行了森林树种分类,并逐一添加新增波段特征确定了各波段的贡献率排名。结果表明：在加入了优选出的均匀性纹理、均值纹理和角二阶矩纹理3种纹理特征后,前4波段和8波段的分类精度比只基于光谱特征时的精度分别高出13.23%和24.63%;利用8波段信息比只利用前4波段在基于光谱特征上的精度高11.88%,在基于光谱+纹理特征上则高23.24%;基于8波段光谱+纹理特征的树种分类精度最高,达到68.74%,新增4波段的贡献率排名为B6>B5>B8>B7,说明新增红边波段对于本次树种分类试验的贡献率最高,能为北方树种识别提供有效帮助。     

北京市政务地理空间信息资源共享服务平台建设实践

从北京市政务地理空间信息资源共享服务平台（以下简称＂平台＂）的建设背景,总体框架,关键技术及创新点、服务现状及成果,经济社会效益等方面介绍了该平台的建设实践。     

GIS技术在房地产市场信息系统中的应用研究

简要介绍了当前国内房地产现状、存在的问题,以及房地产市场信息系统(Real Estate Market Information System,以下简称REMIS)当前所处的发展背景.指出房产GIS在标准和应用上存在的一些问题.对房地产管理和REMIS对GIS的需求情况进行了分析.根据需求分析作者提出了GIS技术在REMIS中的几种应用模式,并指出基于楼盘表为核心的GIS应用模式将是未来房产GIS发展的方向.针对房产GIS应用的不同模式采用的几种关键技术也做了相应描述.     

Effects of Hypoxia on Dopamine Concentration and the Immune Response of White Shrimp （Litopenaeus vannamei）

Effects of hypoxia on the dopamine concentration and the immune response of White Shrimp Litopenaeus vannamei were studied. The results showed that hypoxia had significant effects on the concentration of dopamine (DA) in the haemolymph, haemocyte count, phenoloxidase activity, phagocytic activity of haemocytes and bacteriolytic and antibacterial activity in the haemolvmph (P<0.05). The concentration of the dopamine in haemolymph reached its maximum in the 3.0 and 1.5mgL-1 DO groups at 12h and 6h, and then returned to normal after 24h and 12h, respectively. All immune parameters decreased with the reduction of dissolved oxygen. Total haemocyte count (THC), the hyaline cells and semi-granular cells in the 3.0mgL-1 DO group became stable after 12h,while granular cells did so after 24h, The THC and different haemocyte count (DHC) in the 1.5mgL-1 DO group became stable alter 24h. Phenoloxidase activity and bacteriolytic activity in the 3.0 and 1.5mgL-1 DO groups reached their stable levels after 24 h and 12 h respectively, while phagocytic activity and antibacterial activity became stable after 24 and 12, and 36 and 24 It, respec-tively. It was also indicated that the changes of dopamine concentrations in haemolymph, haemocyte count and phenoloxidase activ-ity were obviously related to the exposure time under hypoxic conditions.     

The selection of alkaline protease-producing yeasts fi＇om marine environments and evaluation of their bioactive peptide production

A total of 400 yeast strains from seawater, sediments, saltern mud, marine fish guts, and marine algae were obtained. The protease activity of the yeast cultures was estimated, after which four strains (HN3.11, N11b, YF04C and HN4.9) capable of secreting extracellular alkaline protease were isolated. The isolated strains were identified as Aureobasidium pullulans, Yarrowia lipolytica, Issatchenkia orientalis and Cryptococcus cf. aureus. The optimal pH of the protease activity produced by strains HN3.11, YF04C, and HN4.9 was 9.0, while that of the protease produced by strain N11b was 10.0. The optimal temperature for protease activity was 45°C for strains HN3.11, N11b, and YF04C, and 50°C for strain HN4.9. After digestion of shrimp (Penaeus vannamei) protein and spirulina (Arthospira platensis) protein with the four crude alkaline proteases, the filtrate from spirulina (Arthrospira platensis) powder digested by the crude alkaline protease of strain HN3.11 was found to have the highest antioxidant activity (61.4%) and the highest angiotensin I converting enzyme (ACE)-inhibitory activities (68.4%). The other filtrates had much lower antioxidant activity and ACE-inhibitory activities.     

基于雷达反射率因子和雷电定位数据的深度学习雷电预报模型

利用卷积神经网络和门控循环单元（Gated recurrent units ）神经网络,基于雷达反射率因子和雷电定位数据开展了雷电预报研究。首先构建了引用注意力机制的基于卷积神经网络和门控循环单元神经网络的深度学习模型（Attention ConvGRU）;然后将雷达反射率因子数据和对应时间段（6 min）的雷电定位数据处理成图像数据后输入深度学习模型,训练出可预报雷电的模型,包括3种模型：单雷电数据模型、单雷达数据模型和雷电〖CD*2〗雷达双数据模型;最后开展了预报试验和定量评估。综合评估表明,本文建立的雷电预报模型综合预报准确率达到96.74%,虚警率35.83%,关键成功指数（Critical Success Index, CSI）为0.2072。个例分析表明,预报模型对于具有明显移动趋势的雷暴过程（A类雷暴）的预报效果优于不具有明显移动趋势的雷暴过程（B类雷暴）,且随着B类雷暴强度减弱模型预报能力逐渐减弱。     

Occurrence and Diversity of Candida Genus in Marine Environments

A total of 317 yeast isolates from seawater,sediments,mud of salterns,guts of marine fishes and marine algae wereobtained.The results of routine identification and molecular characterization showed that six isolates among these marine yeastsbelonged to Candida genus as Candida interrnedia for YA01a,Candida parapsilosis for 3eA2,Candida quercitrusa for JHSb,Can-die rugosa for wl8,Candida zeylanoides for TJY13a,and Candida membranifaciens for W14-3.Isolates YA01a (Candida interme-die),wl8 (Candida rugosa),3eA2 (Candida parapsilosis),and JHSb (Candida quercitrusa) were found producing cell-bound lipase,while isolate W14-3 (Candida membranifaciens) producing riboflavin.These marine yeast Candida spp.Seem to have wide potentialapplications in biotechnology.