STUDY ON MIXED MODEL OF NEURAL NETWORK FOR FARMLAND FLOOD/DROUGHT PREDICTION*

The paper concerns a flood/drought prediction model involving the continuation of time series of a predictand and the physical factors influencing the change of predictand.Attempt is made to construct the model by the neural network scheme for the nonlinear mapping relation based on multi-input and single output.The model is found of steadily higher predictive accuracy by testing the output from one and multiple stepwise predictions against observations and comparing the results to those from a traditional statistical model.     

场对场预报模型及其对降水的预报试验

用经验正交函数对预报因子场和预报对象场分别进行分解，将正交展开的时间系数作为预报因子和预报对象，提出了一种场对场预报模型，该模型考虑了预报系统的时变性，并利用云南５月份的雨量资料进行了预报试验。结果表明，用这种模型作区域性降水预报有一定的优越性。     

An intercomparison of observational precipitation data sets over Northwest India during winter

Development of downscaling models for each calendar month using the data of predictors specifically selected for each calendar month may assists in better capturing the time-varying nature of the predictor-predictand relationships. Such approach will not allow the explicit modelling of the persistence of the predictand (e.g. lag-1 correlation). However, downscaling at an annual time step and subsequent disaggregation to monthly values can explicitly consider the modelling of the persistence of the predictand. This study investigated the potential of annual downscaling of a predictand and subsequent disaggregation of annual values to monthly values, in comparison to the potential of downscaling models separately developed for each calendar month. In the case study, annual and monthly downscaling models were developed for precipitation and evaporation at two stations located in Victoria, Australia. The output of the annual downscaling models was then disaggregated into monthly values using four different methods based on the method of fragments. It was found that the annual to monthly disaggregation methods and monthly downscaling models are able to reproduce the average of monthly observations with relatively higher accuracy in comparison to their ability in reproducing standard deviation, skewness and lag-1 serial correlation. Downscaling models separately developed for each calendar month were able to show relatively smaller root mean square errors for their time series indicating better overall agreement with observations in comparison to their counterpart annual to monthly disaggregation methods. Furthermore, it was found that not only the bias in the output of an annual downscaling model but also the presence of annual totals in the records of observations of a predictand that are very similar in magnitude, but having significantly different sets of fragments, can largely contribute to the poor performance of an annual to monthly disaggregation method.     

多模式集合预报及其降尺度技术在东亚夏季降水预测中的应用

利用动力季节模式输出的匹配域投影技术和多模式集合预报技术对多个国家和城市的站点月平均降水进行预报。预报变量是北京1个站、韩国60个站和曼谷地区8个站点的月平均降水,预报因子是从多个业务动力季节预报模式输出的多个大尺度变量。模式回报数据和站点观测降水数据时段是1983—2003年。降尺度预报降水的技巧是在交叉验证的框架下进行的。匹配域投影方法是设定一个可以活动的窗口在全球范围内大尺度场上进行扫描,寻求与目标站点降水最优化的因子和最相关的区域,目标站点的降水变率就是由该匹配域上大尺度环流场信息决定的。最终预报是用多个降尺度模式预报结果的集合预报(DMME)。多个降尺度模式预报结果的集合预报能显著地提高站点降水的预报技巧。北京站,多个降尺度模式预报结果的集合预报的预报和观测降水的相关系数可以提高到0.71;韩国地区,多个降尺度模式预报结果的集合预报平均技巧提高到0.75;泰国,多个降尺度模式预报结果的集合预报技巧是0.61。     

A numerical method of statistical pattern recognition

A numerical method of statistical pattern recognition is proposed in this paper. Different from the discriminatory analysis method currently used in the mathematic statistics, it is unnecessary to assume that the predictand should be subject to a certain distribution. On the contrary, the statistical relationship between predictand and predictor has been obtained directly with computer according to actual distribution to recognize the category of patterns. Result of forecast has been improved as compared with the usual analytic discriminatory method. The influence of predictor on predictand can be seen clearly from this method and the transparency is good. Therefore, it is better to use the method in very short range forecast for which causality is more obvious.     

双评分准则逐步回归法

阐明了以双评分准则(CSC)作逐步回归的基本原理,通过实例给出了计算方法与步骤。用此法所建模型能同时报好预报对象的数量和类别。     

华南前汛期降水预测模型及其预测试验

将中国华南区域分为东、西2个区,对每个区(8个站)的前汛期(4—6月)平均降水量作自然正交展开(EOF),选取各区累积方差贡献超过75%的前4个主分量作为预报分量。再利用偏最小二乘回归方法结合均生函数方法,提出一种同时考虑预报量自身显著变化周期和前期物理量因子对预报量未来变化影响的预报模型,分别建立东、西区前汛期平均降水量的偏最小二乘回归预报方程。试验结果表明,新的预报模型的预报效果比单纯采用前期物理量因子的逐步回归模型更好,并且其预报能力的提高具有合理的分析依据。      

多元门限回归模型的一种建模方法

本文根据门限自回归模型的基本思想[1]，提出一种多元门限回归模型的建模方法。其特点是充分考虑了预报系统中某些特殊预报因子突变点对预报关系的改变作用。数值实例表明，该模型在模拟和预报精度上比一般线性逐步回归模型有一定程度的提高。     

统计图象识别的一种数值方法

本文提出了统计图象识别的一种数值方法,它不需要假定预报对象服从某种分布,而是根据其实际分布利用微机直接建立预报量和预报因子之间的统计关系,识别图象类别。与常用的解析判别方法相比,它提高了预报效果。在这种方法中,由于能较为清楚地看到预报因子对预报量的影响,透明性好,因此对因果关系较为明显的短时预报更适宜。     

用多元门限回归模型作区域夏季旱涝预测

考虑了气候系统中一些变量突变时对预测关系的改变作用，用多元门限回归模型的建模方法建立长江下游地区夏季旱涝趋势预测模型。拟合效果较理想，用１９９４－１９９５年的独立资料检验，预测结果与实况较为接近。     

Ensemble Prediction of Monsoon Index with a Genetic Neural Network Model

After the consideration of the nonlinear nature changes of monsoon index,and the subjective determination of network structure in traditional artificial neural network prediction modeling,monthly and seasonal monsoon intensity index prediction is studied in this paper by using nonlinear genetic neural network ensemble prediction（GNNEP）modeling.It differs from traditional prediction modeling in the following aspects： （1）Input factors of the GNNEP model of monsoon index were selected from a large quantity of preceding period high correlation factors,such as monthly sea temperature fields,monthly 500-hPa air temperature fields,monthly 200-hPa geopotential height fields,etc.,and they were also highly information-condensed and system dimensionality-reduced by using the empirical orthogonal function（EOF）method,which effectively condensed the useful information of predictors and therefore controlled the size of network structure of the GNNEP model.（2）In the input design of the GNNEP model,a mean generating function（MGF）series of predictand（monsoon index）was added as an input factor;the contrast analysis of results of predic- tion experiments by a physical variable predictor-predictand MGF GNNEP model and a physical variable predictor GNNEP model shows that the incorporation of the periodical variation of predictand（monsoon index）is very effective in improving the prediction of monsoon index.（3）Different from the traditional neural network modeling,the GNNEP modeling is able to objectively determine the network structure of the GNNNEP model,and the model constructed has a better generalization capability.In the case of identical predictors,prediction modeling samples,and independent prediction samples,the prediction accuracy of our GNNEP model combined with the system dimensionality reduction technique of predictors is clearly higher than that of the traditional stepwise regression model using the traditional treatment technique of predictors,suggesting that the GNNEP model opens up a vast range of possibilities for operational weather prediction.     

用多元混合门限回归进行汛期降水量预测试验

在综合考虑预测对象的周期变化和前期外部因子的共同作用后,给出带有周期分量的多元混合门限回归模型。通过青岛汛期降雨量的7年预报试验表明,该模型具有较为稳定的预报能力,值得进一步研究应用。     

Statistical Downscaling of Pattern Projection Using Multi-Model Output Variables as Predictors

A pattern projection downscaling method is employed to predict monthly station precipitation. The predictand is the monthly precipitation at 1 station in China, 60 stations in Korea, and 8 stations in Thailand. The predictors are multiple variables from the output of operational dynamical models. The hindcast datasets span a period of 21 yr from 1983 to 2003. A downscaled prediction is made for each model separately within a leave-one-out cross-validation framework. The pattern projection method uses a moving window, which scans globally, in order to seek the most optimal predictor for each station. The final forecast is the average of the model downscaled precipitation forecasts using the best predictors and is referred to as DMME. It is found that DMME significantly improves the prediction skill by correcting the erroneous signs of the rainfall anomalies in coarse resolution predictions of general circulation models. The correlation coefficient between the prediction of DMME and the observation in Beijing of China reaches 0.71; the skill is improved to 0.75 for Korea and 0.61 for Thailand. The improvement of the prediction skills for the first two cases is attributed to three steps: coupled pattern selection, optimal predictor selection, and multi-model downscaled precipitation ensemble. For Thailand, we use the single-predictor prediction, which results in a lower prediction skill than the other two cases. This study indicates that the large-scale circulation variables, which are predicted by the current operational dynamical models, if selected well, can be used to make skillful predictions of local precipitation by means of appropriate statistical downscaling.     

因子的可预报性和预报模型适用性研究初探

本文从因子场与预报量场之间的整体相关性着手，通过典型相关分析来提取因子信息，用求得的典型变量作为新预报因子，经试用表明，新因子的可预报性比原因子有明显提高。 在预报模型的选择上，本文提出了依据预报模型对历史样本实际预测精度的优劣来衡量预报模型的预测能力，从而选用对历史样本预测精度较高的预报模型作未来预报。     

Statistical downscaling of regional daily precipitation over southeast Australia based on self-organizing maps

This paper presents a novel statistical downscaling method based on a non-linear classification technique known as self-organizing maps (SOMs) and has therefore been named SOM-SD. The relationship between large-scale atmospheric circulation and local-scale surface variable was constructed in a relatively simple and transparent manner. For a specific atmospheric state, an ensemble of possible values was generated for the predictand following the Monte Carlo method. Such a stochastic simulation is essential to explore the uncertainties of climate change in the future through a series of random re-sampling experiments. The novel downscaling method was evaluated by downscaling daily precipitation over Southeast Australia. The large-scale predictors were extracted from the daily NCAR/NCEP reanalysis data, while the predictand was high-resolution gridded daily observed precipitation (1958?C2008) from the Australian Bureau of Meteorology. The results showed that the method works reasonably well across a variety of climatic zones in the study area. Overall, there was no particular zone that stands out as a climatic entity where the downscaling skill in reproducing all statistical indices was consistently lower or higher across seasons than the other zones. The method displayed a high skill in reproducing not only the climatologic statistical properties of the observed precipitation, but also the characteristics of the extreme precipitation events. Furthermore, the model was able to reproduce, to a certain extent, the inter-annual variability of precipitation characteristics.     

多变量均生函数的正交化筛选方案在季节降水预报中的试验研究

用预报量与多变量预报因子的均值生成函数,通过Ｇｒａｍ－Ｓｃｈｍｉｄｔ正交化使均值生成函数彼此正交,同时利用双评分准则确定入选预报模型的均生函数个数,最后建立山西省季节降水的预报模型。试验结果表明,该模型预报山西省季节降水的效果很好,是一种提高短期区域气候预报水平的可行方法。     

CMIP5模式降水订正法及未来30年中国降水预估

借助英国气候研究所（Climate Research Unit, CRU）全球陆地格点分析数据集（CRU TS v4.0）月降水资料和24个国际耦合模式比较计划第五阶段（Coupled Model Intercomparison Project Phase 5, CMIP5）模式历史气候模拟及RCP4.5情景下的降水预估数据,设计了多种回归方案并对模式降水预估偏差进行订正。这些方案包括一元回归、一元对数回归、一元差分回归、一元对数差分回归、多元回归、多元对数回归、多元差分回归、多元对数差分回归和简单移除气候漂移等。2006～2015年中国大陆模式降水预估的订正结果表明,一元回归订正法普遍优于多元回归订正和扣除气候漂移订正法,其中一元对数回归法的效果最好,其降水距平同号率（Anomaly Rate, AR）和降水距平百分率相关系数（Anomaly Percentage Correlation Coefficient, APCC）最高,分别达到69%和0.5;而降水距平相关系数（Anomaly Correlation Coefficient, ACC）最高的是一元对数差分回归法。不同回归订正法所得预估结果的距平同号格点分布显示,一元对数回归法在北方优于南方,而一元差分（年际增量）或对数差分回归法在南方优于北方。这直接导致在中国南方区域（95°E以东,35°N以南）一元对数回归或多元对数回归订正结果的AR、ACC和APCC均低于对应的差分/对数差分回归法,在北方和西部地区则与此相反。因此,模式降水的回归订正方案具有区域性,这可能源于不同区域降水序列统计性质的差异。用区域组合回归订正法,即在南方用一元差分回归订正,其余地区用一元对数回归订正,其降水预估场的AR提高到72%,但ACC和APCC均略有下降,原因是差分回归订正增加了预估降水场的方差。对RCP4.5情景下2016～2045年24个模式集合平均降水预估的组合回归订正结果显示,相对于1976～2005年平均,未来30年降水异常大致呈南北少,中间多的格局,其中长江中下游、江南中西部、西南东北部、华南沿海和海南省等地降水偏少10%～20%,淮河流域、三江源区和台湾省降水偏多10%～40%,西北东部、华北和东北大部降水正常或略偏少。从降水百分率方差看,模式群的离散度（不确定度）呈现东部小,西部大的分布特征,说明模式预估的西北中部和青藏高原西部等降水偏少区的不确定性较大;而河套北部、华北南部和江南东部等地对应于2006～2015年检验期的“盲区”（模式与观测降水距平反号）,其降水预估参考价值可能不大,需要引入他法加以改进。     

Toward the development of prediction models for the primary Caribbean dry season

Summary Two statistical models are created for the Caribbean during its dry season. Canonical correlation analysis (CCA) confirms that there is a robust El Ni?o Southern Oscillation (ENSO) signal in the region during the dry season and that the mode manifests itself as oppositely signed precipitation anomalies over the north and south Caribbean. The south-eastern Caribbean becomes dry in response to a warm event. The first statistical model consequently uses a rainfall index averaged over the south-eastern Caribbean as the predictand. A model which retains an ENSO proxy as one of two predictors shows reasonable skill with hindcast predictions for the region. A second model is created using a Jamaican rainfall index as predictand. Jamaica falls in the transition zone i.e. between the oppositely signed north-south precipitation anomalies characteristic of the ENSO dry season mode. In this case no ENSO related predictor is retained in the final model. Composite analysis of select atmospheric variables for anomalously high and low rainfall years (for the dry season) give an understanding of the dynamics of the Caribbean dry season during phases of the ENSO, particularly those which lead to the creation of the transition zone. Authors’ address: Tannecia S. Stephenson, A. Anthony Chen, Michael A. Taylor, Department of Physics, University of the West Indies, Mona, Jamaica.     

一种神经网络的云图短时预测方法

依据6hT213数值预报产品的资料,采用EOF展开和人工神经网络等方法,对卫星云图短时预报方法进行研究。首先对卫星云图灰度值样本序列进行EOF展开,将提取出来的时间系数作为建模的预报量,以数值预报产品的物理量场作为预报因子,建立人工神经网络预测模型。将预报得到的时间系数与空间特征向量进行时空反演,实现对未来6h云图的预测。预报方法的独立样本试验证明,预测结果与实际云图的主要特征基本吻合,尤其在预测云图的大体分布和发展趋势上得到了较好效果。