基于随机森林特征选择的城市绿化乔木树种分类

城市绿化在改善空气、水和土壤质量,吸收和减少二氧化碳及各种污染物,缓解城市热岛和减少雨水径流等方面发挥着重要作用。及时准确地获取树种信息是城市规划与绿化管理的先决条件,对进一步改善城市生态环境也具有重要意义。基于遥感技术,使用高空间分辨率的WorldView-2卫星影像,采用光谱、纹理、指数以及几何等多种特征相结合的面向对象方法,并通过随机森林进行特征选择,对福州大学旗山校区北部的榕树、杧果、香樟、重阳木、羊蹄甲、垂叶榕以及木棉7种主要绿化乔木进行树种分类。实地验证结果表明：通过特征选择可以减少或规避数据冗余以及休斯效应的产生,该方法可以提高现有同类型树种分类的精度,当淘汰全部特征的20%,利用34个特征（包括15个光谱特征、6个纹理特征、8个指数特征和5个几何特征）进行分类时,总精度最高,可达74.95%,Kappa系数为0.67。其中,光谱平均值的特征重要性最高,而各波段的标准差的重要性较低。WorldView-2卫星影像的4个新增波段,特别是黄光和红边波段及其构建的指数特征重要性较高,也说明这些波段在植被遥感,特别是树种分类中极具应用前景。     

西藏拿顿铜(金)矿床岩矿相学、流体包裹体和地球化学特征与成矿作用研究

拿顿矿床位于多龙矿集区西南部,地处西藏成矿潜力巨大的班公湖-怒江成矿带上,目前该矿床的勘查研究程度非常低,仅在地表圈定出一个铜金矿体,具有很大的找矿前景。本文对拿顿矿床的岩矿相学、流体包裹体及地球化学特征进行了研究。矿床中主要蚀变矿物有高岭土、绢云母、绿帘石、方解石、白云石等,并存在银金矿、辰砂、方铅矿、闪锌矿等特征金属矿物组合。斑晶与胶结物中的流体包裹体均一温度集中在222~360℃之间,主要成分以H2O为主,含有少量CO_2,斑晶中包裹体的盐度为一组高盐度值(18.35%~26.7%NaCleqv.)和一组低盐度值(0.18%~12.85%NaCleqv.),胶结物中包裹体盐度值较低,主要分布在1.6%~12.5%NaCleqv.区间内,是在中低温、低盐度、氧化环境中形成,成矿流体受大气降水的影响较为明显。花岗闪长斑岩中SiO2含量为55.9%~65.67%,平均值为60.52%,Na_2O+K_2O含量为3.54%~5.563%,平均值为4.83%,属于高钾钙碱性系列至钾玄岩系列。A/CNK值为1.63~2.93,平均值为2.6,反映原岩为过铝质岩石。花岗闪长斑岩中稀土总量ΣREE变化范围为29.09×10~(-6)~98.73×10~(-6),整体上具轻微铕负异常(δEu=0.7~1.12),不具铈异常(δCe=0.84~0.99);岩石富集大离子亲石元素(LILE:Rb、K),亏损Nb、Ta、P、Ti等高场强元素(HFSE),具岛弧火山岩的典型特征。矿床脉状黄铁矿中Co/Ni比值为0.05~8.523,平均值为4.219,为热液成因,脉状黄铁矿多与石英脉共生,为后期热液作用阶段的产物。矿床中黄铁矿δ34S值的分布范围为-4.5‰~0.7‰,平均值为-1.95‰,黄铜矿的δ34S值为-7‰~-3.3‰,平均值为-5.43‰,硫同位素δ34S值偏小,且为负值,指示矿床中的硫受到了还原硫的影响。矿床中赋矿岩体为花岗闪长斑岩,发育银金矿、辰砂等特征矿物组合,流体包裹体具有斑岩型铜(金)矿床的中高温流体特征,据此判断矿床成因类型为低硫型浅成低温热液矿床叠加斑岩型铜(金)矿床。根据因子分析,矿床最早期的成岩作用形成早期的中酸性岩体;矿床的成矿作用主要为热液铜金矿化作用,随后发生铅锌矿化、磁铁矿化和钾化等作用,在成矿后期有含碳酸盐热液的侵入。     

藏北高原土壤的温湿特征

通过藏北高原两个站点(D110和安多)土壤温湿特性的分析,表明浅层土壤温度的变化幅度明显的比深层的要大,而且浅层土壤温度受地表随机天气过程的影响较大。浅层(20cm)土壤在未冻结前湿度的变化幅度不但受形成降水的地表随机天气过程的影响,而且受其下层土壤湿度状况的影响。下层土壤湿度越小,浅层(20cm)土壤湿度的变化幅度越大。土壤湿度和土壤温度之间存在着明显的相互关系,土壤的湿度状况能够影响土壤温度变化的幅度和土壤温度变化的趋势。     

人口统计数据的空间转换

在经济和社会研究中,所要研究的区域之上经常没有数据,而这些数据需要由已知区域的数据求得,即统计数据需要空间转换,这就通常涉及到面积内插。本文从GIS的角度研究如何解决人口内插问题,认为面积内插和GIS中的叠加分析是一致的。在传统的面积内插方法的基础上是提出了基于人口真实分布的面积内插方法,并推导出了公式。同时提出了人口密度的递归算法,即把居住区分为人口稀疏地区和人口稠密地区,估计出人口稀疏地区的人口密度,就可以求出人口密集地区的人口密度;再把人口密集区分为新的人口稀疏区和密集区,此过程反复直至求出接近于人口真实分布的人口模型。     

相山居隆庵21号带铀矿床水文地质条件及矿坑涌水量估算

居隆庵菱形断块是相山地区重要铀资源基地,21号带地处居隆庵菱形断块的西部,水文地质条件属简单-中等类型,相对其它矿床更为复杂。笔者采用狭长水平坑道法和水文地质比拟法对该矿床进行矿坑涌水量估算,其钻孔单位涌水量0.1L·(s-1·m-1),预测矿坑涌水量29m3·h-1。研究显示,矿床主要含水带(构造及裂隙)富水性弱且不均,导通性不强;而地下水补给条件好,水压高。在矿床开采过程中应加强监测,及时判断可能出现的涌水量变化情况,保证生产安全。     

Groundwater discharge to the Changjiang River, China, during the drought season of 2006: effects of the extreme drought and the impoundment of the Three Gorges Dam

Groundwater can play an important role in the compensation of runoff reduction due to extreme climate events such as droughts, as well as in response to anthropogenic actions such as the construction of a dam. The increase in ²²⁶Ra specific activity and the runoff from September to December in 2006 is used to estimate the total discharge of groundwater along the mid-lower reaches of the Changjiang River. The total groundwater discharge was found to account for 31% of the increased discharge between Yichang and Datong. The groundwater discharge to lakes (i.e. Dongting Lake) constituted the major contribution of groundwater discharge to the mid-lower reaches of the Changjiang River. More importantly, the second impounding operation of the Three Gorges Dam from 20 September to 27 October 2006 induced a water level decrease in surrounding lakes and rivers, which led to an additional groundwater discharge of 63.3?×?10⁸?m³, accounting for 85% of the total groundwater discharge in the same period. Taken together, these observations indicate that groundwater discharge along the mid-lower reaches plays an important role in maintaining stream flow in the drought season, especially in extreme drought years or in response to human activities.     

浮游植物对海水溶油净化作用的初步研究

本文研究了海水中溶油对骨条藻、角毛藻,三角褐指藻、金藻、甲藻、杜氏盐藻的生长影响。结果表明杜氏盐藻、三角褐指藻、甲藻等在含油海水中9天左右达到最高含量,且在17天左右仍维持最高含量,表明这些藻的耐油性较好。这些耐油藻使海水中石油含量下降3—6倍,表明它们对海水中溶油有净化作用。     

Study of the comprehensive risk analysis of dam-break flooding based on the numerical simulation of flood routing. Part I: model development

Dam-break floods have been of increasing concern to safety engineers and decision makers. The presence of complex terrain in inundation areas multiplies the simulation difficulty of flood routing. In previous studies, representing the flood routing parameters empirically does not reflect the characteristics of flood routing, which strongly influences the accurate assessment of the dam-break consequences. The basis for carrying out dangerous reservoir reinforcement is just engineering safety, without considering the actual situation of downstream areas. In this study, a comprehensive risk analysis of the dam-break flood was implemented based on the numerical simulation of flood routing. First, coupled with the volume of fluid method, a three-dimensional k–? turbulence mathematical model was developed for flood routing in complex inundation areas. Then, based on the flow parameters obtained through computational fluid dynamics modeling, the attribute measure methodology was used for the evaluation of consequences combined with the calculation of the dam-break consequences (loss of life, economic loss, social and environmental influence). Furthermore, a methodology containing the combined weight method and the technique for order performance by similarity to ideal solution method was proposed for risk ranking of dangerous reservoirs due to its logical consideration of scalar values that simultaneously account for both the best and worst alternatives. Finally, a sensitivity analysis was performed to provide information about the stability of risk ranking. The aforementioned model and methodology are applied to a case involving five reservoirs in the Haihe River Basin in China for Part II of this study.