下载文档原格式
基于稳健估计时间序列法的风功率预测朱晓荣;刘艳萍【摘要】Time series method based on robust estimation is introduced to make short term prediction of wind power. Firstly the data are preprocesscd, and then the least squares method and robust estimation method are respectively applied to build an autoregressive integrated moving average model, finally the wind power in the next 30 minutes are forecasted, repeating 10 times. The results show that by using the time series model based on robust estimation to predict wind power, the forecasting error of most points is between 5 percent, except one point of 10. 1 percent. The forecasting error is significantly smaller than conventional time series. It proves robust estimation methods can get a better forecast accuracy when the data have few outliers.%基于稳健估计运用时间序列法对风电场出力进行了短期预测.先对数据进行了预处理,用最小二乘法和稳健估计法分别建立了自回归滑动平均模型.通过模型提前预测了下个30min的风电场出力,总共预测了10次.结果表明,基于稳健估计的时间序列建模进行预测的误差大多数都在5％以内,只有一个点达到10.1％,明显比常规的时间序列建模预测的误差要小.说明稳健估计能在建模数据含有少量异常值时,比常规自回归模型预报精度要高.【期刊名称】《电力系统及其自动化学报》【年(卷),期】2012(024)003【总页数】5页(P107-110,126)【关键词】风电场出力预测;时间序列法;稳健估计;最小二乘法【作者】朱晓荣;刘艳萍【作者单位】华北电力大学电力与电子工程学院,保定071003;华北电力大学电力与电子工程学院,保定071003【正文语种】中文【中图分类】TM614风力的随机性和间歇性不能保证输出平稳的功率，这对电力系统的稳定性以及发电和运行计划的制定带来很多困难。
渤海湾温带风暴潮数值预报模型李大鸣;徐亚男;白玲;解以扬;吴丹朱;何乃光【摘要】To decrease economical losses and environmental pollutions, a numerical model for extratropical storm surge, cogenerated by forces of tide and wind is established, and continuous and fast calculating mode is achieved. Regional nesting method is adopted in this model, open boundary conditions are provided by harmonic analysis, the wind process distributed in time and space is provided by meco-scale models (MM5). The model adopts alternating direction implicit (ADI) for solving equations with explicit and implicit scheme alternately, and the method of local-deepening and water range reducing is to deal with lateral boundary. Based on those conditions, the latest three storm surge processes are simulated by the above-mentioned numerical model, and good results are achieved with the comparisons between the simulated values and observed data, especially on peak value of water elevation. It is proved that the model is of high value to forecast the storm surge for Bohai Bay and can predict surge height in this area.%为减少港区经济损失,降低环境污染,建立了海洋潮波动力和风应力场联合作用的温带风暴潮数值预报模型,实现了连续、快速达到一定精度要求的运行模式.模型采用大小嵌套模型模式,以多分潮调和分析提供海洋水边界条件,以MM5风场计算成果形成时空分布风应力场过程,模型主体网格采用ADI差分格式进行显隐交替计算,浅水变动岸边界采用局部深槽、缩小水域的活动边界处理方法.模型预报并验证了渤海湾海域近期3次风暴潮过程,预报的潮位过程、增水过程与实测值进行比较,潮位过程吻合较好；增水过程在峰值处比较接近,其他各处趋势基本一致；表明该模型在渤海湾海域温带风暴潮预报模拟中具有应用价值,可以用来预报该海区的风暴潮过程.【期刊名称】《天津大学学报》【年(卷),期】2011(044)009【总页数】7页(P840-846)【关键词】风暴潮;预报模型;风应力场;调和分析;渤海湾【作者】李大鸣;徐亚男;白玲;解以扬;吴丹朱;何乃光【作者单位】天津大学建筑工程学院暨港口与海洋工程教育部重点实验室,天津300072;天津大学建筑工程学院暨港口与海洋工程教育部重点实验室,天津300072;天津大学建筑工程学院暨港口与海洋工程教育部重点实验室,天津300072;天津市气象科学研究所,天津300074;天津市气象科学研究所,天津300074;天津市气象科学研究所,天津300074【正文语种】中文【中图分类】P456.7;TV139.2渤海湾海域是中国渤海三大海湾之一，位于渤海西部，海底地形大致自南向北，自岸向海倾斜，属风暴潮灾害的多发区和严重区．其灾害多发生在盛夏台风活动季节和春秋过渡季节[1]．渤海湾中有丰富的石油储藏(大港油田、冀东南堡油田)，湾内有天津新港，每年因为风暴潮灾害经济损失达亿元以上，因此建立适用于渤海湾海域的风暴潮数值预报系统，对减少港区经济损失，降低由溢油扩散及污染源引起的环境污染有重要的经济效益和环境效益．研究表明在温带风暴潮过程中，气压场与波浪场较风应力场在渤海湾海域的影响有限[2-3]，可以在此区域忽略；但由于渤海湾为 3面环陆的半封闭性海湾，岸线较长且地形复杂，温带风暴潮数值预报模型的建立仍具有一定难度．笔者所建立的嵌套模型采用多分潮调和分析方法解决了海域开边界处理问题，提高了渤海湾海域模型空间网格的分辨率，并应用显隐交替的有限差分格式(alternating direction implicit，ADI)对风暴潮控制方程离散求解，采用局部深槽、缩小水域的计算模式处理了动态浅水岸边界，提高了模型计算的稳定性．国内外学者已经发展了多种风暴潮预报模式[4]，早期美国的Jelesnianski等[5]发展了 SPLASH(special program to list amplitudes of surges from hurricane)模式，后在该模式的基础上美国发展了新的 SLOSH(sea，lake and overland surges from hurricanes)模式来预报海上、陆上、湖上的台风风暴潮；Blumberg[6]发展了POM(Princeton ocean model)模式，该模式可以计算小尺度河川的水理运动，也可以模拟大尺度海洋、海岸的水位与流场的变化．在国内，李艳芸等[7]在风暴潮预报模式于渤海海域中的应用研究中，采用COHERENS(coupled hydrodynamic-ecological model for regional and shelf seas)三维多功能大陆架水动力数学模型模拟了热带风暴下渤海的增水过程；于福江等[8]建立了球坐标系下的温带风暴潮模式，对渤海湾一次特大风暴潮过程进行了数值模拟．但连续、快速的温带风暴潮预报模型的研究，特别是对多个例风暴潮过程预报的数学模型研究还需要做大量的工作．笔者自主创建温带风暴潮预报模型，与气象部门合作，已应用于对渤海湾海域长系列天文潮、风暴潮、风暴潮增水的长期、连续、快速预报计算，在大模型网格10 km×10 km，小模型网格1 km×1 km，模拟3,d的72 h风暴潮过程，在现行一般配置的个人计算机上只需运行5 min．笔者选择了近期预报的3次渤海湾风暴潮过程，模拟结果与实际观测进行比较，表明该模型对渤海湾风暴潮过程预报具有一定的实际应用价值．1 数值模型建立的理论基础1.1 水动力数学模型的控制方程温带风暴潮是由大气强迫力(风场、气压)作用于海面造成的海水水位与潮流的剧烈变化．因为渤海湾是平均水深只有18 m的3面环陆的浅海，地形变化复杂，气压对风暴潮过程的影响较小[9]，只考虑风应力场和潮汐作用．模型采用直角坐标系，假定沿水深方向的动水压强分布符合静水压强分布，将三维流动的基本方程沿水深平均积分即可得到沿水深平均的平面二维流动的基本方程[10]．以风应力、水流与海底的摩擦应力[11]为主要影响因子，水动力数学模型控制方程为式中：ξ为增水位；h为平均水深；H为全水深，H＝ξ+h ；u、v分别为x、y方向上流速分量；τx,s、τy,s分别为x和y方向的海面风应力；f为柯氏系数．模型岸边界条件为：v n = 0 (n为边界法线方向)．模型水边界条件为： =0；ξ=ξ(t)(ξ(t)为已知的边界潮位变化过程)．1.2 方程离散求解采用 ADI方法对方程进行离散求解，差分的交错网格为正方形网格，网格线分别平行于x轴和y轴，间距为Δ x = Δy =Δs ．在时间段内，ξ在点(,)ij上，u在点上，隐式求解方程为在点(i , j+ )上，v的显式求解方程为在时间段内，ξ在点(,)ij上，v在点上的隐式求解方程为在点(i + ,j)上，对u显式求解方程为1.3 风场的预报模式模型驱动力风场采用 MM,5模式[12-14]计算成果．MM,5以 NECP资料中的0.5×0.5格距的 GFS资料做背景场，中心点位置选在 41.1,N，118.2,E处．模型网格分辨率9,km，利用探空和地面观测资料对背景场进行修正，从而得到 10,m 风场作为风暴潮的主要驱动力；模型能够预报连续 72 h的逐时风场．将计算得到的经纬度坐标下的风场数据插值到直角坐标系下的模型模拟区域的格点内．模型中风应力计算则采用应用较广泛的公式式中：W为海面10,m风速；aρ为空气密度，取为1.226,kg/m3； DC 按经验取为2.6×10-3．1.4 调和分析方法模型计算区域是整个渤海与部分黄海，开边界为青岛港(35.43,N，119.58,E)到韩国西岸港口HAMPYEONG MAN(35.15,N，126.35,E)的连线，对此边界进行了多分潮调和分析计算．根据验潮站 1 a的潮位实测资料，采用Sa、SSa、Mm、Mf 等 35个分潮[14]，见表 1(部分分潮)．最后 1列为调和分析计算成果．表1 调和分析中采用的部分分潮参数Tab.1 Chon-tide schedule in harmonic analysis分潮代号分潮名称分潮角速度/((°)·h-1)Sa 太阳年分潮 10.041 SSa 太阳半年分潮 10.082 Mm 太阴月分潮 10.544 Mf 太阴半月分潮 11.098Q1 主要太阴椭率日分潮 13.399 O1 主要太阴日分潮 13.943 M2 主要太阴半日分潮 28.984 P1 主要太阳日分潮 14.959 M1 副太阴椭率日分潮14.492 K1 太阴太阳合成日分潮 15.041 J1 二阶太阴椭率日分潮15.585 ς2 主要太阳半日分潮 30.000 2N2 二阶太阴椭率半日分潮27.895µ2 太阴变移半日分潮 27.968 N2 主要太阴椭率半日分潮 28.440周期/平太阳时相对振幅系数8,766.230 1c0.009 4,382.880 110.002 4,661.309 110.003 4,327.859 110.003 4, 326.868 112.949 4, 325.819 115.692 4, 312.421 100.000 4, 324.066 116.067 4, 324.841 110.064 4, 323.934 120.000 4, 323.098 111.152 4, 312.000 135.775 4, 312.905 112.494 4, 312.872 113.011 4, 312.658 119.063各分潮的调和常数即振幅与迟角的展开公式[15]为式中：0a为基准面的平均海平面高度；0V u+ 为分潮的天文初相角； jH、 jg为分潮振幅与迟角；j、m为分潮序号与总数；jσ为分潮角频率；jf为平均振幅的订正系数；t为时间．1.5 动态浅水岸边界处理ADI差分计算模式要求整个计算域应保持在水深以下，对浅水岸边界的露滩、淹没变化应是连续、稳定过程．本文采用局部深槽、缩小水域的活动边界处理方法，当全水深接近0.1 m时，在浅水网格区全水深保持为 H 10 = 0 .1m ，保持流量、流速不变，流量为变化后的水域宽度 SB为2 风暴潮数值预报模型的应用及结果分析图1 模型计算区域范围及嵌套模式示意Fig.1 Computed domain and setup of nesting domain模型采用嵌套网格模式如图1所示，网格剖分为正方形网格[16]，其中第 1套网格构成的大模型空间步长为10 km，模型范围涵盖整个渤海及部分黄海海域，模型计算域跨度为117°38'47"E至126°32'38"E，35°18'5"N至40°50'43"N，计算结果包括部分黄海及渤海的潮流与潮位过程，为第2套网格提供海洋开边界条件；第 2套网格构成的小模型空间步长为 1,km覆盖整个渤海湾，岸边界网格更加精细化，为实现岸边界的浅水动态处理提供了条件，模型网格见图2．大模型网格总数为78×61，计算时间步长 60,s；小模型网格总数121×161，计算时间步长10,s，其中图2(b)的标记点为渤海湾内的主要港口和地区，包括塘沽、曹妃甸、黄骅港等，模型以塘沽为验证点，计算了3次典型的风暴潮过程，验证结果表明，所建风暴潮数值预报模型可应用于渤海湾地区的风暴潮预警预报系统.图2 黄渤海区域与渤海湾网格划分示意Fig.2 Mesh grid of Yellow-Bohai Sea and Bohai Bay2.1 调和分析结果验证由 2002年青岛港实测潮位资料计算出 35个分潮的调和常数，以此分别计算选定的3个不同时段的青岛港潮位过程，将计算结果与实测资料对比验证如图 3所示．对比曲线表明，应用调和分析方法具有可行性．韩国西岸港口HAMPYEONG MAN的潮位资料采用潮位软件提供，模型以青岛港潮位过程为主，考虑韩国西岸港口潮位过程修正，确定大模型计算域的开边界条件．图3 青岛港潮位调和分析验证Fig.3 Comparison and analysis of tidal level in Qingdao Harbor2.2 风场模式计算结果分析图4 2007-08-12 14:00黄渤海海域与渤海湾风速分布Fig.4 Distribution of wind speed of Yellow-Bohai Sea and Bohai Bay at 2 pm Aug 12th, 2007渤海湾为3面环陆的内海，风暴潮过程受风场影响最为显著，尤其受到东风与东北风的作用后海面水位抬高显著．模型预报的渤海湾 3次典型风暴潮过程分别为070812次、090415次和090509次，风场在渤海湾的分布主要呈东北偏东风形势．如图 4所示，渤海湾 070812次风暴潮受自身海域气旋影响较小，而受黄海海域气旋影响显著，从而导致潮位超过警戒水位；影响 090415次风暴潮过程的风场是东北大风，风速最大达到 21.29,m/s；090509次风暴潮过程直接受当日的7级偏东大风影响，导致塘沽验潮站的风暴潮潮位超过警戒水位．警戒水位预设为4.7,m．2.3 数值模拟结果图5 风暴潮潮位过程曲线Fig.5 Tidal level process curves of storm surge将已建立的风暴潮数值预报模型应用到渤海湾，调试验证了渤海湾 070812次风暴潮、同时对风暴潮090415次、090509次进行了预报检验．图 5与图 6给出了计算潮位与实测潮位过程和计算增水值与实测增水值的曲线对比验证．从潮位的变化趋势看，070812次风暴潮计算潮位与实测潮位拟合一般相关系数为 0.93，同时 090415次、090509次预报的风暴潮潮位与实测潮位符合程度较高相关系数分别为0.96和 0.97(见表 2)；从增水过程变化趋势看，对 3次风暴增水的峰值能够较好地模拟出来．070812次风暴潮实测最高潮位 4.83,m，增水 0.86,m，而模型预报结果显示塘沽最高潮位达 4.96,m，增水 1.38,m，与实测值较为接近；090415次风暴潮受冷空气与气旋影响，塘沽验潮站测得该日6∶00发生4.94,m大潮，增水为 1.84,m，模型提前 2日计算得到 4月 15日6∶00风暴潮最高潮位为 5.08,m，计算增水达到1.94,m；090509次风暴潮受南下冷空气影响，风力最高达 7级，实际观测塘沽站最高潮位 4.95,m，增水1.27,m，模型提前2日计算结果及误差分析如表2所示，模型模拟潮位峰值的平均相对误差为 0.024，增水峰值的平均相对误差也在允许的范围内．上述结果表明，建立的风暴潮数值预报模型能够对渤海湾海区风暴潮进行较为准确的预报，可以用来对该地区风暴潮进行预报．但是，该模型在模拟增水峰值出现的时间上与实测值存在误差，其原因是多方面的：在对天文潮进行非线性模拟时，由于缺失水边界实测资料采用调和分析方法，虽然考虑了 35个分潮的作用，但以青岛港的潮位值来代替整个水边界的潮位值，可能导致峰值与相位的偏差；在本文所建立的风暴潮预报模型中，近岸地形采用的是20世纪90年代资料，这与实际近岸地形分布可能存在一定偏差，从而造成计算模拟误差，但是随着预报经验的增加和资料的不断收集完善，可望减小预报误差．图6 风暴潮增水验证曲线Fig.6 Tidal rising process curves of storm surge表2 塘沽站风暴潮潮位及增水值误差分析Tab.2 Error analysis on tidal leveland tidal rising of storm surge at Tanggu station风暴潮次号比较项目观测值/m 计算值/m 相对误差相关系数070812 最大增水 0.86 0.930.081 0.93最高潮位 4.83 5.04 0.043 090415 最大增水 1.84 1.940.054 0.96最高潮位 4.94 5.08 0.028 090509 最大增水 1.27 1.380.087 0.97最高潮位 4.95 4.96 0.0023 结论(1)渤海湾温带风暴潮数值预报模型，采用大小嵌套模型模式，以多分潮调和分析提供海洋边界条件，以MM5风场计算成果形成时空分布风应力场过程，模型主体网格采用 ADI差分格式进行显隐交替计算，实现了连续、快速达到一定精度要求的风暴潮预报运行模式，为渤海湾温带风暴潮数值预报提供了可靠的研究手段．(2)渤海湾属浅海类型海湾，岸滩地形变化复杂，本文中提出浅水变动岸边界的计算模式，在 ADI差分格式连续计算中，采用局部深槽、缩小水域的活动边界处理方法，增加了模型计算的稳定性，提高了模型模拟风暴潮在岸滩附近增水计算的能力．(3)本研究建立的海洋潮波动力和风应力场联合作用的温带风暴潮数值预报模型，预报并验证了渤海湾海域近期 3次风暴潮过程，预报潮位过程、增水过程与实测值进行比较，潮位过程吻合较好；增水过程在峰值处比较接近，平均相对误差较小，其他各处趋势基本一致，表明该模型在渤海湾海域温带风暴潮预报模拟中具有一定的应用价值，可以用来预报该海区的风暴潮过程．【相关文献】［1］李大鸣，徐亚男，宋双霞，等. 波浪辐射应力在渤海湾海域对风暴潮影响的研究[J]. 水动力学研究与进展A辑，2010，25(3)：374-382.Li Daming，Xu Yanan，Song Shuangxia，et al. The study on the effect of wave-radiation stress on storm surge in Bohai Bay[J]. Chinese Journal of Hydrodynamics，Ser A，2010，25(3)：374-382(in Chinese).［2］ Andrey P，Eppel D P，Kapitza H. Interaction of waves，currents and tides，and wave-energy impact on the beach area of Sylt Island[J]. Ocean Dynamics，2009，59(3)：451-461.［3］林祥，尹宝树，侯一筠，等. 辐射应力在黄河三角洲近岸波浪和潮汐风暴潮相互作用中的影响[J]. 海洋与湖沼，2002，33(6)：615-621.Lin Xiang，Yin Baoshu，Hou Yijun，et al. Effects of radiation stress in the interaction of coupled wave-tidesurge in the coastal area of Huanghe Delta[J]. Oceanologia Etlimnologia Sinica，2002，33(6)：615-621(in Chinese). ［4］ Kim K O，Yamashita T. Storm surge simulation using wind-wave-surge coupling model[J]. Journal of Oceanography，2008，64：621-630.［5］ Jelesnianski C P，Chen J，Shaffer W A. SLOSH：Sea，Lake，and Overland Surges from Hurricanes[R].National Weather Service，USA，1992.［6］ Blumberg A F，Mellor G L. A description of a threedimensional coastal ocean circulation model[C]//Three-Dimension Coastal Ocean Models. Washington，USA，1987：1-16.［7］李艳芸，李绍武. 风暴潮预报模式在渤海海域中的应用研究[J]. 海洋技术，2006，25(1)：101-106.Li Yanyun，Li Shaowu. Application research of a storm surge prediction model in Bohai Sea[J]. Ocean Technology，2006，25(1)：101-106(in Chinese).［8］于福江，王喜年，宋珊，等. 渤海“9216”特大风暴潮过程的数值模拟[J]. 海洋预报，2000，17(4)：9-15.Yu Fujiang，Wang Xinian，Song Shan，et al. The numerical simulation of storm surge in Bohai Sea caused by tropical storm POLLY[J]. Marine Forecast，2000，17(4)：9-15(in Chinese).［9］陈士荫，顾家龙. 海岸动力学[M]. 北京：人民交通出版社，1988.Chen Shiyin，Gu Jialong. Coastal Dynamics[M].Beijing：People’s Communication Press，1988(in Chinese).［10］ Jones J E，Davies A M. Storm surge computations for the west coast of Britain using a finite element model(TELEMAC)[J]. Ocean Dynamics，2008，58(5/6)：337-363. ［11］ Lee D. Bottom shear stress under wave-current interaction[J]. Journal ofHydrodynamics，Ser B，2008，20(1)：88-95.［12］ Lee S M，Princevac M，Mitsutomi S，et al. MM5 simulations for air quality modeling：An application to a coastal area with complex terrain[J]. Atmospheric Environment，2009，43(2)：447-457.［13］ Heo K Y，Lee J W，Ha K J，et al. Simulation of atmospheric states for a storm surge on the west coast of Korea：Model comparison between MM5，WRF and COAMPS[J]. Nat Harzards，2009，51(1)：151-162.［14］ Ivanov S，Palamarchuk J，Pyshniak D. Upscale feedbacks through microphysics fields at nesting domains of the MM5 model[J]. Atmospheric Research，2009，94(1)：726-735.［15］陈宗镛. 潮汐学[M]. 北京：科学出版社，1980.Chen Zongyong. Tidal Science[M]. Beijing：Science Press，1980(in Chinese).［16］于福江，张占海. 一个东海嵌套网格台风暴潮数值预报模式的研制与应用[J]. 海洋学报，2002，24(4)：23-33.Yu Fujiang，Zhang Zhanhai. Implementation and application of a nested numerical typhoon storm surge forecast model in the East China Sea[J]. Acta Oceanologica Sinica，2002，24(4)：23-33(in Chinese).。
风电专业术语文件编码（008-TTIG-UTITD-GKBTT-PUUTI-WYTUI-8256）风电专业术语风能：wind energy 空气流动所具有的能量。
?[wind ?en?d?i]风能资源：wind energy resources 大气沿地球表面流动而产生的动能资源。
空气的标准状态：standard atmospheric state 空气的标准状态是指空气压力为101 325Pa，温度为15℃(或绝对，空气密度m 3 时的空气状态。
风速：wind speed 空间特定点的风速为该点空气在单位时间内所流过的距离。
平均风速：average wind speed 给定时间内瞬时风速的平均值。
年平均风速：annual average wind speed 时间间隔为一整年的瞬时风速的平均值。
最大风速：maximum wind speed 10分钟平均风速的最大值。
极大风速：extreme wind speed 瞬时风速的最大值。
阵风：gust 超过平均风速的突然和短暂的风速变化。
年际变化：inter-annual variation 以30年为基数发生的变化。
风速年际变化是从第1年到第30年的年平均风速变化。
[风速或风功率密度]：年变化：annual variation 以年为基数发生的变化。
风速(或风功率变化)年变化是从1月到12月的月平均风速(或风功率密度)变化。
[风速或风功率密度]：日变化：diurnal variation 以日为基数发生的变化。
月或年的风速(或风功率密度)日变化是求出一个月或一年内，每日同一钟点风速(或风功率密度)的月平均值或年平均值，得到0点到23点的风速(或风功率密度)变化。
风切变：wind shear 风速在垂直于风向平面内的变化。
风切变指数：wind shear exponent 用于描述风速剖面线形状的幂定律指数。
风速廓线：wind speed profile, wind shear law 又称“风切变律”，风速随离地面高度变化的数学表达式。
北半球平流层大气环流转型的基本气候特征张灵;李维京;陈丽娟【期刊名称】《应用气象学报》【年(卷),期】2011(022)004【摘要】利用1948-2009年NCEP/NCAR逐日高度场和风场再分析资料探讨了平流层各主要层次上环流转型的年际、年代际时空演变特征.结果表明:北半球平流层冬季环流转为夏季环流的过程是高层环流转型早,低层环流转型晚,但在各层次上环流转型早晚存在着区域性差异.自新地岛到西伯利亚北部地区的环流转型最早,且该区域与北半球环流平均转型时间的年际以及年代际特征最相近.北半球平流层环流转型的气候平均时间早于东亚热带季风爆发时间,从而可能成为季风预测的前兆信号.分析还得到平流层各主要层次环流转型时间具有明显的年代际特征,环流转型时间呈现由偏晚到偏早、又从偏早到偏晚的变化特征,只是年代际转折年份在不同区域、不同层次存在差异.此外,平流层环流转型时间普遍存在准2年、准3～6年、准9～12年以及准21～24年的周期,可能与气候系统其他成员有密切联系.%The basic climatic features of stratospheric circulation in Northern Hemisphere demonstrate different forms in winter and summer. In winter, the cold cyclone system and westerly winds prevail in high latitudes, while in summer the situation is the opposite. In terms of inversion of geopotential height gradient and zonal wind direction, a transition date index (TDD indicating the change dates from summer to winter circulations in the stratosphere in Northern Hemisphere is defined by using NCEP/NCAR reanalysis daily data. Some statistic methods such as linear tendency,wavelet analysis, binomial coefficient smooth and Mann-Kendall are applied to analyze the inter-annual and inter-decadal features of the transition dates at all main levels in the stratosphere. Results indicate that in the stratosphere, with the height rising, the transition date becomes earlier and the summer circulation lasts longer. For instance, the earliest circulation transition in the stratosphere occurs at the height of 10 hPa and 20 hPa, and it shifts to 30 hPa in a short period. However, it takes longer for the transition to shift from 30 hPa to 50 hPa than that from 10 hPa to 30 hPa. Which takes almost one month. The average onset date of the South China Sea Summer Monsoon (SCSSM) is one of the earliest dates in Asia Summer Monsoon (ASM) system and it is much later than the transition dates in stratosphere. Therefore, TDI can be used as a pre-signal for monitoring and predicting ASM. Furthermore, there exists an obvious regional difference in the circulation transition, among which the transition dates at each level in Siberia is the earliest and that is relatively later in Bering Sea and Greenland. The inter-annual and inter-decadal features of the circulation transition dates in Northern Hemisphere and the aforementioned three different regions are quite apparent, turning from late to early and then to late again in the past 62 years. Particularly the circulation transition date in Northern Hemisphere and in Siberia shares some similarities in inter-annual and inter-decadal variations, for example, the time variation shows significant fluctuations, and both have a transition peak in 1975. The transition dates in Bering Sea and Greenland also have the similar features, for example, the time fluctuation is relatively small.Moreover, circulation transition dates vary with the height and region, but they all have a quasi-2-year, a quasi-3-to-6-year, a quasi-9-to-12-year or a quasi-21-to-24-year cycle which may have close connections with other members of the climate system.【总页数】10页(P411-420)【作者】张灵;李维京;陈丽娟【作者单位】兰州大学大气科学学院,兰州730000;国家气候中心中国气象局气候研究开放实验室,北京100081;国家气候中心中国气象局气候研究开放实验室,北京100081;国家气候中心中国气象局气候研究开放实验室,北京100081【正文语种】中文【相关文献】1.北半球大气环流能量循环的气候特征 [J], 李新新;管兆勇;李明刚2.ENSO 年冬季北半球平流层大气环流异常特征分析 [J], 李琳;李崇银;谭言科3.北半球平流层各高度场谐波分析的若干基本事实 [J], 瞿章;吕世华;郑光;林俊峰;邱海龙4.大气环流变化时南北半球、东西半球对流层与平流层研究 [J], 田荣湘5.青藏高原OLR的气候特征及其对北半球大气环流的影响 [J], 李栋梁;章基嘉;吴洪宝因版权原因，仅展示原文概要，查看原文内容请购买。
nCounter miRNA Expression Assay User Manual MAN-C0009-08About this ManualThis manual describes the methods for miRNA Sample Preparation and miRNA CodeSet Hybridization. For instructions on post-hybridization processing and data analysis, please see the instrument-specific user manuals (nCounter Pro Analysis System User Manual (MAN-10147), nCounter Analysis System User Manual for MAX/FLEX systems (MAN-C0035), nCounter SPRINT Profiler User Manual (MAN-10017))and the Gene Expression Data Analysis Guidelines (MAN-C0011).Changes in this Revision (MAN-C0009-08)•Updates to text for clarity and accuracy.•Updates to list of required materials.•Increased volume of ligation master mix (step 7 on page 8) to ensure sufficient volume of viscous solution.•Adjusted recommended hybridization time to 16-24 hours to align with best practices.Table of ContentsAbout this Manual (2)Changes in this Revision (2)Introduction (3)Workflow (4)Materials and Equipment (4)Materials Supplied by NanoString (4)Additional Materials Required (5)Equipment (5)Thermal Cycler Guidelines (5)Sample Guidelines and Recommendations (6)miRNA Sample Preparation Protocol (7)miRNA CodeSet Hybridization Setup Protocol (9)Technical Support (11)MAN-C0009-08 nCounter miRNA Expression Assay User ManualIntroductionNanoString® patented molecular barcodes provide a true digital detection technology capable of highly multiplexed analysis. The nCounter® miRNA Expression Assay is designed to provide an ultra-sensitive and reproducible method to detect miRNAs without the use of reverse transcription or amplification. The assay is run on total RNA isolated from any source, including formalin-fixed paraffin embedded (FFPE) samples, and allows for detection across all biological levels of expression.This assay involves sample processing where unique oligonucleotide tags are annealed and subsequently ligated with miRNAs of interest via a target specific bridge oligo (Figure 1, top). Sequence specificity between each miRNA and its appropriate tag is ensured by careful, stepwise control of annealing and ligation temperatures. This sample processing allows the short miRNAs to be detected with great specificity and sensitivity using NanoString Codeset chemistry. CodeSet chemistry consists of a Reporter CodeSet and a Capture ProbeSet that hybridize to the specific targets of interest (Figure 1, bottom).Excess probes are removed, and target-probe complexes are immobilized and aligned on the cartridge.Their barcodes are then counted on an nCounter system.Figure 1. miRNA-specific preparation to utilize nCounter technology downstream.The tabulated barcodes from the nCounter system can be analyzed using NanoString’s nSolver™ Analysis Software or other analysis programs, such as the ROSALIND™ cloud platform.nCounter miRNA gene expression panels are sold in increments of 12 reactions. Master Kits (for MAX, FLEX, or Pro Analysis Systems) or SPRINT Reagents and Cartridges (for SPRINT Profilers) are also required and sold separately (see Table 2).nCounter miRNA Expression Assay User Manual MAN-C0009-08WorkflownCounter technology allows a simple workflow of less than 45 minutes hands-on time and streamlined data analysis in under 24 hours to profile miRNAs of interest.Table 1. Workflow for the nCounter miRNA Assay.Day 1(2 hr total) miRNA Sample Preparation Protocol 30 minutes miRNA CodeSet Hybridization Setup Protocol 5 minutesDay 2(up to 6 hr total) Set up either the Prep Station run or SPRINT run 5 minutes Move Cartridge to Digital Analyzer for Data Collection (Pro/MAX/FLEX only) 5 minutesMaterials and EquipmentMaterials Supplied by NanoStringTable 2. NanoString-provided materials required to run the nCounter miRNA Assay.miRNA CodeSet:- Human v3: Catalog # CSO-MIR3-12 - Mouse v1.5: Catalog # CSO-MMIR15-12 - Rat v1.5: Catalog # CSO-RMIR15-12miRNA Reporter CodeSetmiRNA Capture ProbeSet-80°C-80°CnCounter miRNA Sample Prep Kit: - Human: Catalog # Hu-MIRTAG-12 - Mouse: Catalog # Mu-MIRTAG-12 - Rat: Catalog # Rn-MIRTAG-12Annealing BuffermiRNA Tag ReagentPEGLigation BufferLigaseLigation Clean-up EnzymemiRNA Assay Controls-20°C-20°C-20°C-20°C-20°C-20°C-80°CnCounter Master Kit (for MAX/FLEX/Pro) -Catalog # NAA-AKIT-012nCounter Sample CartridgePrep PlatePrep Pack, including Hybridization Buffer-20°C4°C15–25°CnCounter SPRINT Reagent Pack (for SPRINT)-Catalog # SPRINT-REAG-KIT nCounter SPRINT Reagent CnCounter SPRINT Reagent A, B, andHybridization Buffer4°C15–25°CnCounter SPRINT Cartridge (for SPRINT)-Catalog # SPRINT-CAR-1.0nCounter SPRINT Cartridge -20°CMAN-C0009-08 nCounter miRNA Expression Assay User ManualAdditional Materials RequiredTable 3. Additional materials required (not provided by NanoString). Pipettes for 0.5–10 μL, 2–20 μL, 20–200 μL, and 200-1000 μLVariousMulti-channel pipette for 20 μL (optional)RNase-free pipette tips with aerosol barriers1.5-mL microcentrifuge tubes0.2-mL strip tubes and caps, nuclease-free (SPRINT users only; these areprovided in Master Kits for MAX/FLEX/Pro users) Disposable glovesMolecular biology-grade nuclease-free water EquipmentTable 4. Required equipment to run the nCounter miRNA Assay.Spectrophotometer or fluorometerNanoDrop Technologies® spectrophotometer, or ThermoFisher® Qubit TM fluorometer Picofuge or mini-centrifuge with strip tube adaptorStratagene® or equivalent Thermal cycler with a programmable heated lid(see Thermal Cycler Guidelines, below)Various NanoString nCounter Pro, MAX, or FLEX AnalysisSystem or SPRINT Profiler NanoString Technologies, IncThermal Cycler GuidelinesPlease note that a thermal cycler with a heated lid is required for this protocol. NanoString recommends a model with a programmable heated lid, to avoid high temperatures that cause tubes to melt or deform. •NanoString recommends a thermal cycler with a heated lid that can adjust throughout the protocol.The heated lid should be set to 5°C greater than the current incubation temperature at any moment.•Otherwise, program the heated lid to be 5°C greater than the maximum temperature reached in the protocol. The heated lid should not exceed 110°C.nCounter miRNA Expression Assay User Manual MAN-C0009-08 Sample Guidelines and RecommendationsThe nCounter miRNA Expression Assay requires purified total RNA as input material. NanoString recommends approximately 100 ng of total RNA to generate robust signal for most tissue and cell isolates.Total RNA purified from any cell or tissue type may be used, including formalin-fixed, paraffin-embedded (FFPE) material. A variety of kits are available to extract RNA from FFPE (such as Qiagen® miRNeasy®, Catalog # 217504). Ensure that the selected kit preserves the small RNAs.For plasma, serum, or other cell-free sample-derived RNA, please see the Tech Note for nCounter miRNA Expression Analysis in Plasma and Serum Samples (MK1432)for important sample preparation considerations. Unpurified lysates may not be used with the nCounter miRNA Expression assay, as the denaturants in the homogenization buffer will inhibit the sample preparation reaction.The quality of the purified RNA is critically important for the nCounter miRNA assay. Residual contaminants left over from lysis and RNA extraction can impact assay performance by inhibiting the enzymatic ligation and purification steps. Typical lysis or extraction contaminants that can inhibit the assay include:•Guanidinium Isothiocyanate (lysis buffer)•Guanidinium HCl (initial wash buffer)•Phenol (organic extraction)•Ethanol (secondary wash buffer) Purified RNA quality can be evaluated via a spectrophotometer by measuring absorbance at 230 nm (A230), 260 nm (A260) and 280 nm (A280). Significant absorbance at 230 nm, or a low A260/A230 ratio, can indicate contamination with organic compounds such as phenol or guanidinium. Extra washes with a secondary wash buffer or ethanol can help to minimize carry-through. Be sure to remove residual secondary wash buffer prior to elution/resuspension. Significant absorbance at 280 nm, or a low A260/A280 ratio, can indicate contamination with protein. Such contamination may lead to an overestimation of the RNA concentration, resulting in a lower-than-anticipated signal in the assay.NanoString recommends a 260/280 ratio of 1.9 or greater and a 260/230 ratio of 1.8 or greater for optimal results. Please note that for plasma, serum, or other cell-free sample-derived RNA, the mRNA content will be too low to obtain accurate absorbance measurements.IMPORTANT:•At very low RNA concentrations (under ~10 ng/μL), the A260/A230 ratio may be unreliable as an indicator of contamination, due to limited nucleic acid absorbance at 260 nm. For tissue and cell derived RNA,NanoString recommends preparing samples with a concentration of > 33 ng/μL, allowing 100 ng oftotal RNA to be added to the sample preparation reaction in the available 3 μL volume.•Some RNA extraction protocols suggest that better yield can be achieved by re-eluting the column with the initial eluate. NanoString does not recommend this, as the extra elution can generatesignificant carry-through of guanidinium and organic contamination. If re-elution must be performed,it should be preceded by at least 2 additional column washes with the secondary ethanol-based washbuffer (for a total of 4 secondary washes).•Ethanol is not evident spectrophotometrically. Excess ethanol can be eliminated by a one-minute post-wash centrifugation in a clean collection tube, as is suggested in most kit protocols, and/or by air-drying the filter for 5 minutes.MAN-C0009-08 nCounter miRNA Expression Assay User Manual miRNA Sample Preparation ProtocolnCounter miRNA assays require purified total RNA as input material. See the Sample Guidelines and Recommendations on page 6 for additional information on sample input considerations.All experiments should be designed in sets of 12 samples. NanoString reagents are supplied in 12-reaction aliquots. The protocol below is for one set of 12 samples.1.Program a thermal cycler according to the protocols in Table 5.2.Prepare a 1:500 dilution of the miRNA Assay Controls: combine 499 μL of nuclease-free water with1 μL of the miRNA Assay Controls in a sterile microcentrifuge tube. Mix by vortexing and briefly spindown. Store on ice.NOTE: Assay control RNA included in the nCounter miRNA Sample Preparation Kit allows the user tomonitor the ligation efficiency and specificity through each step of the reaction.3.Prepare an annealing master mix: combine 13 μL of Annealing Buffer, 26 μL of nCounter miRNA TagReagent and 6.5 μL of the 1:500 miRNA Assay Controls dilution prepared in Step 2. Mix well bypipetting up and down.4.Aliquot 3.5 μL of the annealing master mix into each tube of a 0.2-mL strip tube.5.Add 3 μL of RNA sample (recommended concentration ≥ 33 ng/μL) to each tube. Cap tubes and flickgently to mix. Spin down.nCounter miRNA Expression Assay User Manual MAN-C0009-086.Place strip in thermal cycler and initiate the Annealing Protocol (Table 5).bine 22.5 μL PEG and 15 μL Ligation Buffer to prepare a 15X ligation master mix. Mix well bypipetting up and down.NOTE: PEG is viscous and should be pipetted slowly to ensure accurate transfer of volume into themix.8.Following completion of the Annealing Protocol, when the thermal cycler has reached 48°C, add 2.5μL of the ligation master mix to each tube. (Do not turn off the thermal cycler; the block must be at48°C in Steps 9 and 10). Flick tubes gently to mix and spin down.9.Return tubes to 48°C thermal cycler, close lid, and incubate at 48°C for 5 min.IMPORTANT: For Step 10, do not remove tubes from the thermal cycler. Maintaining the temperatureof the tubes at 48°C is critical for optimal assay performance.10.Open thermal cycler, carefully remove caps from tubes (leaving strip in place in the heat block) andadd 1 μL of Ligase directly to the bottom of each tube while tubes remain at 48°C. Check the pipettetip to be sure all the Ligase was added to the reaction. There is no need to mix.NOTE: To keep track of Ligase addition to sequential samples, it can be helpful to line up 12 tips infront of thermal cycler, discarding each tip after use.11.Immediately after addition of Ligase to the final tube, recap tubes (leaving tubes in heat block), closethermal cycler, and initiate the Ligation Protocol (Table 5).12.After completion of the Ligation Protocol, add 1 μL Ligation Clean-Up Enzyme to each reaction. Thetubes can be removed from the heat block for this step. Flick tubes gently to mix, then spin down.13.Return tubes to thermal cycler and initiate the Purification Protocol (Table 5).14.After completion of the Purification Protocol, add 40 μL nuclease-free water to each sample. Mix welland spin down.NOTE: At this point, purified sample preparation reactions may be stored at –20°C for several weeks.15.Proceed with the miRNA CodeSet Hybridization Setup Protocol (page 9).NOTE: Be sure to denature your samples as part of the CodeSet Hybridization Protocol (Step 6.b. ofthe following protocol).MAN-C0009-08 nCounter miRNA Expression Assay User ManualmiRNA CodeSet Hybridization Setup ProtocolReporter CodeSet and Capture ProbeSet Handling Instructions:• During setup, do not vortex or pipette vigorously to mix. Instead, gently flick or invert thetubes.• To spin down contents of tubes, a picofuge or mini-centrifuge is recommended. If using acentrifuge, spin at <3000x g for <10 seconds. Do not “pulse” spin as it will cause the centrifuge to go to maximum speed and may spin the probes out of solution.The order of addition of components is important; follow the protocol exactly.1. Pre-heat a thermal cycler to 65°C with a heated lid at 70°C; set the time interval to “infinite”. Do notset the thermal cycler to ramp down to 4°C at the end of the incubation time.NOTE : A thermal cycler with a heated lid is required for this protocol. NanoString recommends athermal cycler with a programmable heated lid. See Thermal Cycler Guidelines on page 5.2. Remove Reporter CodeSet and Capture ProbeSet reagents from the -80°C freezer and thaw on ice,shielded from light. Once thawed, invert several times to mix well and briefly spin down reagents.IMPORTANT : After it has thawed, inspect the tube of Reporter CodeSet to make sure no coloredprecipitate is present. If you see a colored precipitate, heat the entire tube to 75°C for 10 minutes andcool to room temperature before using.3. Create a hybridization master mix by adding 130 μL Hybridization Buffer to the tube containing 130μL Reporter CodeSet (Table 6). Do not add the Capture ProbeSet to the master mix . Table 6. Hybridization master mix for one nCounter assay (12 reactions + 2 reactions of dead volume).4. Flick or invert the hybridization master mix tube repeatedly to mix, then briefly spin down.5. Label a strip tube . If necessary, cut strip in half to fit in a picofuge with strip tube adaptor, and labelboth halves. For MAX/FLEX/Pro users, use the strip tubes provided with the nCounter Master Kits,ensuring that the notch is positioned between tubes 1-2 and 8-9.6. Prepare the hybridization reactions (Table 7) using a new pipette tip at every step:a. Add 20 μL of hybridization master mix to each tube of the prepared strip tube.miRNA Reporter CodeSet130 (in tube) 10 Hybridization Buffer130 10 Total Volume of Master Mix 260 20nCounter miRNA Expression Assay User Manual MAN-C0009-08b.Denature samples from the miRNA Sample Preparation Protocol at 85°C for 5 minutes, thenquick-cool on ice.c.Add 5 μL of each sample from the miRNA Sample Preparation Protocol to their respectivetubes of the strip tube.d.A dd 5 μL of Capture ProbeSet to each tube.NOTE: Minimizing the time between addition of the Capture ProbeSet and placement of thereaction at 65°C will increase the sensitivity of the assay.e.Cap the strip tube tightly and mix by inverting or flicking to ensure complete mixing.f.Spin briefly and immediately place the strip tube in a preheated 65°C thermal cycler.Table 7. Hybridization reaction contents.Hybridization Master Mix 20Sample from the miRNA Sample Preparation Protocol 5miRNA Capture ProbeSet 5Total Reaction Volume 307.Incubate hybridization reactions for 16-30 hours. Hybridizations should be left at 65°C until readyfor processing, although maximum hybridization time should not exceed 30 hours.NOTE: NanoString recommends 20-22 hours of hybridization time for this assay.8.Once removed from the thermal cycler, proceed immediately to post-hybridization processing withthe nCounter Pro Analysis System User Manual (MAN-10147), nCounter Analysis System UserManual for MAX/FLEX systems (MAN-C0035), or nCounter SPRINT Profiler User Manual (MAN-10017). Do not store hybridizations at 4°C.MAN-C0009-08 nCounter miRNA Expression Assay User ManualFor more information, please visit NanoString Technologies, Inc.530 Fairview Avenue North Seattle, Washington 98109T (888) 358-6266F (206) 378-6288 *******************Sales ContactsUnited States: ***********************EMEA: ***************************Asia Pacific & Japan: *************************Other Regions: *******************Technical SupportFor technical support, please contact **********************.Intellectual Property RightsThis nCounter miRNA Expression Assay User Manual and its contents are the property of NanoString Technologies, Inc. (“NanoString”), and are intended for the use of NanoString customers solely in connection with their operation of an nCounter Analysis System or SPRINT Profiler. The nCounter Analysis System and SPRINT Profiler (including both software and hardware components) and this User Manual and any other documentation provided to you by NanoString in connection therewith are subject to patents, copyright, trade secret rights, and other intellectual property rights owned by or licensed to NanoString. No part of the software or hardware may be reproduced, transmitted, transcribed, stored in a retrieval system, or translated into other languages without the prior written consent of NanoString. For a list of applicable patents, see /company/patents.TrademarksNanoString, NanoString Technologies, the NanoString logo, nCounter, and nSolver are trademarks or registered trademarks of NanoString Technologies, Inc., in the United States and/or other countries. All other trademarks and/or service marks not owned by NanoString that appear in this document are the property of their respective owners.© 2022 NanoString Technologies, Inc. All rights reserved.For Research Use Only. Not for use in diagnostic procedures.。
博士学位论文波浪对表层海流以及Ekman层风能输入的影响作者姓名：回贞立指导教师：徐永生研究员中国科学院海洋研究所学位类别：理学博士学科专业：物理海洋学研究所：中国科学院海洋研究所2017年6月The impact of waves on ocean surface currents and wind energy input to the Ekman layerA Dissertation Submitted toUniversity of Chinese Academy of SciencesIn partial fulfillment of the requirementFor the degree ofDoctor of PhilosophyByZhenli HuiDissertation Supervisor: Professor Xu YongshengInstitute of Oceanology, Chinese Academy of SciencesJune 2017摘 要由卫星数据反演的表层海流包含两部分：由散射计风场数据反演的Ekman 流以及高度计海表面高度数据（SSH）反演的地转流。
然而经典的Ekman模型并没有考虑波浪的影响，通过将波浪引起的Coriolis-Stokes力考虑在内，研究了波浪（主要是Stokes漂流）对卫星反演的海表面流的影响，并将产品与OSCAR （Ocean Surface Current Analyses Real-time）表层流和拉格朗日漂流浮标实测流进行了对比，结果显示，考虑了Stokes漂流的影响后得到的表层流产品与漂流浮标实测流的结果最为吻合，特别是在南大洋海域（40°S~65°S），与不考虑Stokes 漂流的表层流相比，90% （91%）的纬向（经向）流得到了改善。
对于纬向流速，与漂流浮标的相关系数（均方根误差）从0.78（13 cm/s）增加（减小）到0.81（10.99 cm/s），对于经向流速，相关系数（均方根误差）从0.76（10.87 cm/s）增加（减小）到0.79（10.09 cm/s），该发现说明波浪的确对大洋环流有重要的影响，在全球大洋环流的数值模拟中，应该被考虑在内。
第30卷 第5期2023年5月仪器仪表用户INSTRUMENTATIONVol.302023 No.5风电惯量响应及一次调频控制策略研究综述赵 亚，张 君（南京工程学院，南京 210000）摘 要：风电并网比例不断提高降低了电力系统频率稳定性，风力发电应具备惯量响应与一次调频能力。
从风电渗透率提高对电力系统频率稳定性的影响与风电惯量响应及一次调频控制策略两方面进行了分析综述，主要分析对比了虚拟惯性控制、有功-频率下垂控制、减载控制、储能调频及综合协调控制等不同方案，展望了需要进一步深入研究的关键问题。
关键词：风电；惯量响应；一次调频；频率稳定中图分类号：TM614 文献标志码：AOverview of Wind Power Inertia Response and PrimaryFrequency Modulation Control StrategyZhao Ya ，Zhang Jun（Nanjing Institute of T echnology , Nanjing,210000, China ）Abstract：The increasing proportion of wind power connected to the grid reduces the frequency stability of the power system, and wind power generation should have inertia response and primary frequency regulation capabilities. The influence of wind power penetration rate increase on the frequency stability of power system, wind power inertia response and primary frequency regulation control strategy are analyzed and reviewed. This paper mainly analyzes and compares different schemes such as virtual inertial con-trol, active-frequency droop control, load shedding control, energy storage frequency regulation and comprehensive coordinated control. Key issues requiring further in-depth study are looked at.Key words：wind power ；inertia response ；primary frequency modulation ；frequency stability收稿日期：2023-03-01作者简介：赵亚（1999-），男，江苏徐州人，硕士研究生，研究方向：风电一次调频。
