共查询到20条相似文献，搜索用时 62 毫秒
1.
Most real-world time series have some degree of nonstationarity due to external perturbations of the observed system; external driving forces are the essential reason that leads to the nonstationarity of dynamics system. In this paper, the authors present a novel technique in which the authors incorporate external forces to predict nonstationary time series. To test the effect, the authors also examined two prediction experiments with an ideal time series from a logistic map and a proxy climate dataset for the past millennium. The preliminary results show that the resulting algorithm has better predictive ability than the one that does not consider the external forces. 相似文献
2.
气候系统具有非平稳特征,根本原因在于其外强迫随时间发生改变,因此外部驱动力的分析对于理解气候系统的动力学特征至关重要,而如何有效提取系统外部驱动信息是一个亟待解决的前沿科学问题。最近几年,在生物神经学领域中应用的一种提取非平稳信号中外强迫信息的方法——慢特征分析法(Slow Feature Analysis,SFA),在气象领域中也得到了初步成功的尝试,结果显示出此方法对气候系统的外强迫信息分析及有关动力学机制的探究有较好的应用前景。本文主要介绍SFA方法的理论思想及实施步骤,并通过一个理想的非平稳时间序列检验其提取外强迫信息的能力,结果证明在衰减的Logistic模型中,可利用SFA算法提取出模型中的外强迫,且与真实外强迫的相关系数可达0.99;此外,还介绍将该方法应用于Arosa臭氧时间序列,分析其提取的外强迫信息的动力学特征;并介绍了在气候时间序列建模中引入外强迫因子的预测效果。 相似文献
3.
基于EMD 和集合预报技术的气候预测方法 总被引：3，自引：0，他引：3
气候系统是典型的非平稳性系统,然而对于气候观测数据的处理通常是在时间序列平稳的假定下完成的,比如气温和降水的多步预报,这通常会导致预报准确度较低。为改进该缺陷,首先将非平稳数据序列分解成平稳的、多尺度特征的本征模态函数分量(IMF),再使用数值集合预报与逐步回归分析相结合的方式对每一个IMF 分量构建不同的预报模型,最后线性拟合成预报结果。通过Visual Studio 2008 开发平台使用上述方法建立了一个短期气候预报系统,采用广西区88 个气象站1957—2005 年的2 月距平气温数据进行实际验证。结果表明,相对于普通预测和单一预测方法,加入了EMD 和集合预报技术的方法在仅用历史资料进行多步预测的情况下,对于气候的变化趋势以及突发性气候具有更好的预报能力。 相似文献
4.
中国北方地区旱涝的年代际预测分析研究 总被引：7，自引：8，他引：7
基于状态空间重构理论和嵌入定理,给出场时间序列预测模型的建立思路。与单点时间序列预测分析方法相比,场时间序列预测分析方法的优点在于,在寻找吸引子上某个相点的最邻近点及其映象以建立预测模型时,不局限于它自身的时间序列,而是在区域内所有相点的时间序列所构成的整个吸引子上寻找。这样,在一定程度上改进单点时间序列的“遍历性”,以提高预测精度。在此基础上,利用中国北方地区534年旱涝等级资料,对中国北方几个区域年代际尺度的旱涝变化及其极端旱(涝)出现频率进行预测试验分析。 相似文献
5.
Shuoben Bi Shengjie Bi Xuan Chen Han Ji Ying Lu 《Asia-Pacific Journal of Atmospheric Sciences》2018,54(4):611-622
Observed climate data are processed under the assumption that their time series are stationary, as in multi-step temperature and precipitation prediction, which usually leads to low prediction accuracy. If a climate system model is based on a single prediction model, the prediction results contain significant uncertainty. In order to overcome this drawback, this study uses a method that integrates ensemble prediction and a stepwise regression model based on a mean-valued generation function. In addition, it utilizes empirical mode decomposition (EMD), which is a new method of handling time series. First, a non-stationary time series is decomposed into a series of intrinsic mode functions (IMFs), which are stationary and multi-scale. Then, a different prediction model is constructed for each component of the IMF using numerical ensemble prediction combined with stepwise regression analysis. Finally, the results are fit to a linear regression model, and a short-term climate prediction system is established using the Visual Studio development platform. The model is validated using temperature data from February 1957 to 2005 from 88 weather stations in Guangxi, China. The results show that compared to single-model prediction methods, the EMD and ensemble prediction model is more effective for forecasting climate change and abrupt climate shifts when using historical data for multi-step prediction. 相似文献
6.
The Prediction of Non-stationary Climate Series Based on Empirical Mode Decomposition 总被引：1，自引：0，他引：1
This paper proposes a new approach which we refer to as ``segregated
prediction" to predict climate time series which are nonstationary. This
approach is based on the empirical mode decomposition method (EMD), which
can decompose a time signal into a finite and usually small number of basic
oscillatory components. To test the capabilities of this approach, some
prediction experiments are carried out for several climate time series. The
experimental results show that this approach can decompose the
nonstationarity of the climate time series and segregate nonlinear
interactions between the different mode components, which thereby is able to
improve prediction accuracy of these original climate time series. 相似文献
7.
8.
气候系统模式输出结果是当前开展气候预测业务的重要参考依据之一,如何提高气候系统模式输出结果的可信度是改进气候业务预测能力的关键之一。利用1999—2010年NCEP CFSv2模式每日四次预测未来45天的回算数据,分析了集合样本数对模式预测能力的影响。分析结果表明,模式对月平均500 hPa位势高度的预测技巧在热带地区较高,而中高纬度地区较低;模式对500 hPa位势高度时间异常的预测能力优于空间异常。无论是空间异常还是时间异常,随着模式超前时间的增加,预测技巧均逐渐降低,但是在不同区域和不同月份,模式预测技巧随超前时间的变化存在差异。此外,模式预测技巧存在非常大的年际变率。增加集合样本数,对不同月份和不同起报时间预测技巧的稳定度和预测技巧值均有明显正效果,特别是对亚洲中纬度地区改善度较大。增加集合样本数也可以在一定程度上降低模式预测技巧年际变率。集合样本数增加对于500 hPa位势高度空间异常的改进优于时间异常。 相似文献
9.
The Evolutionary Modeling and Short-range Climatic Prediction for Meteorological Element Time Series 
下载免费PDF全文
下载免费PDF全文 1. Introduction Let us suppose that the meteorological element series is the set of solution by integrating a perfect cli- matic numerical model with certain initial conditions, boundary conditions etc., thus it is also the concen- trated expression of nonlinear interaction between all climatic factors (including itself) in the model. Be- cause of limited understanding the mechanism of cli- matic system changes, the unsolved problems are not less than the solved ones in the climatic numerical … 相似文献
10.
传统的统计方法难以很好的对气候系统这一集非线性、非平稳性为一身的多层次系统进行处理。因此集层次化处理和平稳化处理的集合正交经验模态分解技术(EEMD)的提出,为解决上述问题提供了有效的途径。本文选取黄河中上游24个气象观测站的逐月降水资料,结合组合预报和集合预报思路,基于EEMD建立了统计预报模型。其中对降水序列中的高频部分进行了二次平稳化处理,实现对2008—2013年6—8月的降水预报,并用预报评分检测预报效果。结果表明:EEMD模型对黄河中上游夏季降水有着较强的预报能力,在该区域与气候模式和传统的统计方法相比具有更高的精度和更好的应用前景。 相似文献
11.
动力气候模式是目前国际上开展气候预测的主要工具。经过 8年多的研制、发展和业务化过程 ,国家气候中心已建立起第一代动力气候模式预测业务系统 ,并以此为平台 ,形成了一套包括月、季节到年际时间尺度的动力模式预测业务。 2 0年历史回报试验和 1年多的试验性业务运行结果表明 ,该系统对东亚区域的季节预测具有较好的预测能力 ,其预测结果已经在实际业务中得到了应用 ,并成为我国短期气候预测业务的重要参考依据。该文是对该动力模式系统性能的介绍 ,也是对国家“九五”重中之重课题的加强课题“短期气候预测综合动力模式预测系统业务化”专题的总结汇报。 相似文献
12.
基于1981-2016年江苏省不同区域一季稻产量序列,计算站点尺度的气温、降水、日照适宜度以及综合气候适宜度,在此基础上构建基于气候适宜指数的作物产量预报模型,开展不同空间尺度的一季稻产量精细化预报。同时,以各主产地市、县一季稻种植面积百分比为权重,加权集成省级产量,开展基于不同空间尺度一季稻产量序列的大区域尺度产量预测。结果表明:(1)江苏省不同区域一季稻气象产量与不同时段气候适宜指数之间存在较高的相关性,基于气候适宜指数的预报方法适用于江苏省不同区域一季稻单产预报。(2)2012-2016年省级尺度模型预报平均准确率高于97.5%,主产地市、县模型平均预报准确率低于省级尺度预报模型,主产县预报准确率年际间波动较大,表明预报区域越小,预报的难度提升。(3)基于气候适宜指数模型的江苏省级、主产地市集成,主产县集成模型预报准确率大部在95%以上,整体上看主产县集成优于主产地市集成,主产地市集成优于省级尺度模型。由此,开展地市级和县级尺度的精细化产量预报可提升省级尺度预报准确率,同时提高县级作物产量预报能力。 相似文献
13.
With the high-speed development of numerical weather prediction, since the later 1980’s, the prediction of short-range climate anomalies has attracted worldwide meteorologists’ attention. What the so called short-range re-fers to the time scale from one month to one season or more. In dealing with the problem of short-range climate pre-diction, two points are needed noticing: one is the basic research to explore or investigate the mechanism of variability of the slow varying components which mainly include internal dynamics of extratropics, external forcings and tropical dynamics, and the other is the modeling efforts to simulate the process of the long-term evolution of the signal which include the improvement of model quality, stochastic prediction and the air-sea-coupled model (Miyakoda et al.,1986). Previous researches on the numerical prediction of short-term climate anomalies are mostly concentrated in the analysis of variables with global spatial scale, especially the global general atmospheric circulation analysis.As to the simulation or prediction of regional short-term climate anomalies, there exist many difficulties and problems. Though some meteorologists are devoting themself to this field, up to now, they have not reached satisfac-tory results. As a primary effort, by using the 2-level general atmospheric circulation model developed in the Institute of Atmospheric Physics, Chinese Academy of Sciences (IAP-AGCM) (Zeng et al., 1989), and taking the year of 1985 as a case, a numerical simulation of regional short-term climate change is completed. We pay high attention to the predictand of anomalous summer rainfall in the Yangtze River and Yellow River valleys, especially its month-to-month variation. 相似文献
14.
基于国家气候中心气候系统模式1.1版本（BCC_CSM1.1m）的历史回报数据,利用时间相关系数和均方根误差等确定性技巧评分,对西伯利亚高压、阿留申低压、东亚冬季风3种东亚地区冬季典型环流系统的预报技巧进行检验评估,并通过时间序列分析和空间相关系数等方法,分析东亚地区冬季典型环流系统的可预报性来源。结果表明:由于模式对热带海洋和北太平洋海平面气压的预测偏差小、对欧亚大陆的预测偏差大,模式对阿留申低压、东亚冬季风的预测技巧高于西伯利亚高压。进一步分析表明:厄尔尼诺和南方涛动（ENSO）是阿留申低压和东亚冬季风的重要可预报性来源,而土壤温度是西伯利亚高压的重要可预报性来源,并受ENSO调制。此外,东亚冬季风的预报技巧也受到西伯利亚高压预报技巧的制约。 相似文献
15.
16.
Summary To meet the challenge of developing a comprehensive weather and climate prediction model which can give realistic scenarios for many time scales, more computer power than is currently available will be needed. One possibility for alleviating this shortcoming is to increase the integration timestep. We propose and test several methods which may prove useful. One procedure is an expansion of the model dependent variables in a Taylor series. Application of this method to simple models indicates acceptable increases in timestep by a factor of five. A multi-level approach which is less complex to apply gives comparable results and is more successful when high accuracy is desired. To bypass the limiting constraint of the Courant-Friedrichs-Lewy (CFL) condition on gravity waves, an approach is suggested in which the prediction model is represented in its normal modes and the high frequency modes are balanced while the low frequency modes are predicted. Experiments with this procedure are described and in combination with the multi-level integration technique show substantial increases in integration timestep for acceptable integration results, both on the forecast and climate scale. Experiments are now underway applying this process to the NCAR/CCM3, a state-of-the-art model.With 10 Figures 相似文献
17.
用统计方法作月、季尺度的短期气候乃至年际尺度的长期气候预测是当前气候预测业务的主要依据,在短时间内这种情况仍然不可能彻底改变。虽然数值预报模式的预测能力达到了7 d的时效,不过要积分到月、季尺度并实现短期气候预测还面临着重重困难。其根本原因是气候系统的混沌分量和非线性/非平稳性等因素在起作用。而现有气候预测的统计方法(主要包括经验统计、数理统计和物理统计等方法)的数学基础却忽略了这些特点,这是因为以现有的科学水平人们不得不假设时间序列是线性和平稳的。实际气候观测序列普遍具有层次性、非线性和非平稳性,这给建立预测方法带来了极大困难。文中构建了一个新的预测模型,即首先利用经验模态分解(em-pirical mode decomposition,EMD)方法将气候序列作平稳化处理,得到一系列平稳分量-本征模函数(intrinsic modefunction,IMF);其次,利用均生函数(mean generate function,MGF)模型获得各分量的初次预测值;最后,在最优子集回归(optimal subset regression,OSR)模型的基础上,通过直接或逐步拟合一部分预测值,构建两种预测方案达到提高预测能力的目的。典型气候序列的预测试验结果表明,具有平稳化的IMF分量,尤其是特征IMF分量有较高的可预测性,它对原序列趋势的预测有重要指示意义。大力开展气候系统机理和气候层次的研究,并建立相应的气候模式是未来发展趋势。该文是这方面的一个初步尝试,相信该模型能为气候预测(评估)开辟一条新的有效途径。 相似文献
18.
Eduardo Zorita Anders Moberg Lotta Leijonhufvud Rob Wilson Rudolf Brázdil Petr Dobrovolný Jürg Luterbacher Reinhard Böhm Christian Pfister Dirk Riemann Rüdiger Glaser Johan Söderberg Fidel González-Rouco 《Climatic change》2010,101(1-2):143-168
Two European temperature reconstructions for the past half-millennium, January-to-April air temperature for Stockholm (Sweden) and seasonal temperature for a Central European region, both derived from the analysis of documentary sources and long instrumental records, are compared with the output of climate simulations with the model ECHO-G. The analysis is complemented by comparisons with the long (early)-instrumental record of Central England Temperature (CET). Both approaches to study past climates (simulations and reconstructions) are burdened with uncertainties. The main objective of this comparative analysis is to identify robust features and weaknesses in each method which may help to improve models and reconstruction methods. The results indicate a general agreement between simulations obtained with temporally changing external forcings and the reconstructed Stockholm and CET records for the multi-centennial temperature trend over the recent centuries, which is not reproduced in a control simulation. This trend is likely due to the long-term change in external forcing. Additionally, the Stockholm reconstruction and the CET record also show a clear multi-decadal warm episode peaking around AD 1730, which is absent in the simulations. Neither the reconstruction uncertainties nor the model internal climate variability can easily explain this difference. Regarding the interannual variability, the Stockholm series displays, in some periods, higher amplitudes than the simulations but these differences are within the statistical uncertainty and further decrease if output from a regional model driven by the global model is used. The long-term trend of the CET series agrees less well with the simulations. The reconstructed temperature displays, for all seasons, a smaller difference between the present climate and past centuries than is seen in the simulations. Possible reasons for these differences may be related to a limitation of the traditional ‘indexing’ technique for converting documentary evidence to temperature values to capture long-term climate changes, because the documents often reflect temperatures relative to the contemporary authors’ own perception of what constituted ‘normal’ conditions. By contrast, the amplitude of the simulated and reconstructed inter-annual variability agrees rather well. 相似文献
19.
An analytic solution of an energy balance model (EBM) is presented which can beused as a recursive filter for time series analysis. It is shown that the EBM can reproduce the solution of a coupled atmosphere-ocean general circulation model (AOGCM) experiment. Contrary to the AOGCM, the EBM easily allows for variations in climate sensitivity to satisfy the full range of uncertainty concerned with this parameter. The recursive filter is applied to two natural and two anthropogenic forcing mechanisms which are expressed in terms of heating rate anomaly time series: volcanism, solar activity, greenhouse gases (GHG), and anthropogenic tropospheric aerosols. Thus, we obtain modelled global mean temperature variations as a response to the different forcings and with respect to the uncertainty in the forcing approximations and climate sensitivity. In addition, it is shown that the observed (ENSO-corrected) global mean temperature time series within the period from 1866 to 1997 can be explained by the external forcings which have been considered and an additional white noise forcing. In this way we are able to separate different signals and compare them. As a result, global anthropogenic climate change due to GHG forcing can be detected at a high level of significance without considering spatial patterns of climate change but including natural forcing, which is usually not done. Furthermore, it is shown that solar forcing alone does not lead to significantclimate change, whereas solar and volcanic forcing together lead to a significant natural climate change signal. Anthropogenic climate change due to GHG forcing may partly be masked by anthropogenic aerosol cooling. 相似文献
20.
利用集合经验模态分解(Ensemble Empirical Mode Decomposition,EEMD)有效地分解了全球平均表面温度(Global Mean Surface Temperature,GMST)时间序列,得到其不同尺度的、不同特征的子序列(Intrinsic Mode Function,IMF)。在此基础上,利用在预测长期、复杂、非线性变化的时间序列上具有显著优势的滑动自回归机器学习(Autoregressive Integrated Moving Average,ARIMA)模型和长短期记忆网络(Long Short-Term Memory,LSTM)模型开展GMST年际信号预测研究。结果表明:深度学习模型LSTM能很好地拟合并预测了长程相关性强的子序列(第2~6个IMF),而代表GMST年际尺度变化的IMF1则在一定程度上受到太平洋大西洋多重气候信号的影响和调制,因此进一步将3个气候指数作为预报前兆因子加入预测模型来更准确地预测IMF1的时间演变。通过利用多套GMST数据的对比,最终选定了考虑实时ENSO信息的LSTM(ENSO)模型来提前预测年际GMST信号,并预测2020年将有较大概率会成为史上最热的年份之一。 相似文献