CN104992057A - Quasi-ensemble-variation based mixed data assimilation method - Google Patents
Quasi-ensemble-variation based mixed data assimilation method Download PDFInfo
- Publication number
- CN104992057A CN104992057A CN201510353774.7A CN201510353774A CN104992057A CN 104992057 A CN104992057 A CN 104992057A CN 201510353774 A CN201510353774 A CN 201510353774A CN 104992057 A CN104992057 A CN 104992057A
- Authority
- CN
- China
- Prior art keywords
- forecast
- ensemble
- quasi
- assimilation
- hour
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
- 238000000034 method Methods 0.000 title claims abstract description 35
- 238000004422 calculation algorithm Methods 0.000 claims abstract description 7
- 238000004458 analytical method Methods 0.000 claims description 11
- 239000011159 matrix material Substances 0.000 claims description 9
- 230000003068 static effect Effects 0.000 claims description 8
- 230000001419 dependent effect Effects 0.000 claims description 3
- 238000006467 substitution reaction Methods 0.000 claims 1
- 238000004364 calculation method Methods 0.000 abstract description 10
- 230000000694 effects Effects 0.000 description 5
- 238000004800 variational method Methods 0.000 description 3
- 238000005516 engineering process Methods 0.000 description 2
- 239000000203 mixture Substances 0.000 description 2
- 206010003591 Ataxia Diseases 0.000 description 1
- 206010010947 Coordination abnormal Diseases 0.000 description 1
- 238000009795 derivation Methods 0.000 description 1
- 208000016290 incoordination Diseases 0.000 description 1
- 230000004807 localization Effects 0.000 description 1
- XLYOFNOQVPJJNP-UHFFFAOYSA-N water Chemical compound O XLYOFNOQVPJJNP-UHFFFAOYSA-N 0.000 description 1
Landscapes
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
本发明公开了一种基于准集合-变分的混合资料同化方法，包括如下步骤：选取预报时刻相邻的过去连续一个月历史预报数据中，每6小时保存的12小时和24小时的预报数据，将该数据作为准集合预报样本；计算同一时刻24小时预报与12小时预报的差异，得到准集合预报误差；计算准集合预报误差的均值，将该均值以及准集合预报误差代入无偏估计公式，得到无偏估计；将无偏估计代入准集合-变分同化算法中，进行混合同化。本发明将历史预报误差计算得到准集合背景误差，用于准集合-变分混合资料同化。该准集合背景误差通过相邻的历史预报结果产生，而不需要真实的集合预报，有效降低了集合预报带来的计算量，提高了业务中资料同化和预报的效率。
The invention discloses a mixed data assimilation method based on quasi-ensemble-variation, comprising the following steps: selecting 12-hour and 24-hour forecast data stored every 6 hours from the historical forecast data of the past continuous month adjacent to the forecast time , take the data as a quasi-ensemble forecast sample; calculate the difference between the 24-hour forecast and the 12-hour forecast at the same time to obtain the quasi-ensemble forecast error; calculate the mean value of the quasi-ensemble forecast error, and substitute the mean value and the quasi-ensemble forecast error into the unbiased estimation formula , get an unbiased estimate; Substitute the unbiased estimate into the quasi-set-variational assimilation algorithm to perform mixed assimilation. The invention calculates the historical forecast error to obtain the quasi-ensemble background error, which is used for the assimilation of the quasi-ensemble-variation mixed data. The quasi-ensemble background error is generated by adjacent historical forecast results without real ensemble forecast, which effectively reduces the amount of calculation brought by ensemble forecast and improves the efficiency of data assimilation and forecast in operations.
Description
技术领域technical field
本发明涉及一种基于准集合-变分的混合资料同化方法，属于数值天气预报中的资料同化技术领域。The invention relates to a mixed data assimilation method based on quasi-ensemble-variation, and belongs to the technical field of data assimilation in numerical weather forecasting.
背景技术Background technique
数值天气预报质量由数值预报模式和模式初始场共同决定。目前，数值预报的模式结构及物理过程方案已趋于完善，可以较准确地描述和模拟真实天气系统的演变。所以提高数值天气预报准确性的任务更多的落向如何改善模式初始场——数值天气预报对初始条件的精确性的要求也越来越高。随着软件和硬件技术和观测系统的发展，全球气象观测网的不断升级，观测时间密度和空间分布的不断增加，观测资料类型和数量不断增加，如何有效的利用这些资料为数值天气预报提供更准确的初始场，是我们面临的更进一步提高数值预报水平的关键问题。The quality of numerical weather prediction is jointly determined by the numerical prediction model and the initial field of the model. At present, the model structure and physical process scheme of numerical forecasting have tended to be perfected, which can more accurately describe and simulate the evolution of real weather systems. Therefore, the task of improving the accuracy of numerical weather prediction falls more on how to improve the initial field of the model—the numerical weather prediction has higher and higher requirements for the accuracy of the initial conditions. With the development of software and hardware technology and observation systems, the continuous upgrading of the global meteorological observation network, the continuous increase of observation time density and spatial distribution, and the continuous increase of observation data types and quantities, how to effectively use these data to provide better numerical weather prediction Accurate initial field is the key problem we face to further improve the level of numerical prediction.
目前，资料同化已经被广泛用于融合各种观测信息来为数值模式产生更合理的初始场。研究和业务中使用较多的主要有三维变分同化法、四维变分同化法、集合卡尔曼滤波同化法，以及目前受到学者们较多关注的集合-变分结合的混合同化方法。变分法和集合卡尔曼滤波法结合的混合同化方案，综合了集合卡尔曼滤波法背景场误差协方差可以随天气形势演变的优点，又利用了变分法已经形成一套有效的、成熟的技术方案，被认为是资料同化的主要发展方向。At present, data assimilation has been widely used to fuse various observational information to generate more reasonable initial fields for numerical models. The three-dimensional variational assimilation method, the four-dimensional variational assimilation method, the ensemble Kalman filter assimilation method, and the hybrid assimilation method of ensemble-variation combination that are currently receiving more attention from scholars are mainly used in research and business. The hybrid assimilation scheme combining the variational method and the ensemble Kalman filter method combines the advantages of the ensemble Kalman filter method that the background field error covariance can evolve with the weather situation, and utilizes the variational method to form a set of effective and mature Technical solutions are considered to be the main development direction of data assimilation.
混合同化方案中，背景场误差协方差用一组集合预报表示的集合背景场误差协方差与变分同化中静态的背景场误差协方差相结合。混合同化方案缓解了集合方案不满秩、变量不协调问题，也改善了变分方案模型化背景场误差协方差各向同性和匀质性、无法依天气形势而变的问题，许多学者也对混合同化方案进行了大量的研究测试，大多数的研究结果都表明：混合同化方法的预报效果优于单纯的变分方法，且在集合成员较少的情况下，它也能达到与集合卡尔曼滤波同化法相似的效果。资料同化的目的就是寻找一个最优的分析场使得目标函数最小，集合-变分混合同化方法的目标函数可表示为：In the hybrid assimilation scheme, the background field error covariance is combined with the ensemble background field error covariance represented by a set of ensemble forecasts and the static background field error covariance in variational assimilation. The mixed assimilation scheme alleviates the problems of dissatisfaction of rank and incoordination of variables in the set scheme, and also improves the isotropy and homogeneity of the error covariance and homogeneity of the background field modeled by the variational scheme, which cannot change according to the weather situation. A large number of research tests have been carried out on the assimilation scheme, and most of the research results show that the forecasting effect of the hybrid assimilation method is better than that of the simple variational method, and it can also achieve the same effect as the ensemble Kalman filter in the case of fewer ensemble members. Assimilation has a similar effect. The purpose of data assimilation is to find an optimal analysis field to minimize the objective function. The objective function of the ensemble-variation hybrid assimilation method can be expressed as:
其中，J为目标函数，同化目地就是不断修正δx，使目标函数J最小。δx1＝x-xb，δx1为传统三维变分同化时的增量，x为分析场，xb为背景场，B为静态背景误差协方差矩阵，β1为静态协方差的权重系数，A为变量相关矩阵，起到局地化的作用，α为集合扩展控制变量，β2为流依赖协方差的权重系数，H为线性化观测算子，R为观测误差协方差矩阵，d＝y-H(xb)为观测增量，其中，y为观测场，H为线性化观测算子， Among them, J is the objective function, and the purpose of assimilation is to continuously modify δx to minimize the objective function J. δx 1 =xx b , δx 1 is the increment of traditional three-dimensional variational assimilation, x is the analysis field, x b is the background field, B is the static background error covariance matrix, β 1 is the weight coefficient of the static covariance, A is the variable correlation matrix, which plays the role of localization, α is the set expansion control variable, β 2 is the weight coefficient of the flow-dependent covariance, H is the linearized observation operator, R is the observation error covariance matrix, d=yH (x b ) is the observation increment, where y is the observation field, H is the linearized observation operator,
对于常规混合同化，式(1)中为集合预报误差的无偏估计：For conventional mixed assimilation, in formula (1) is an unbiased estimate of the ensemble forecast error:
xi为第i个集合预报成员，N为集合预报成员数，为集合预报平均。从式(2)可以看出，混合在引入集合预报误差协方差的时候，需要集合成员的计算，而且集合成员如果太少还会带来集合预报误差协方差不满秩、变量不协调问题虽然混合同化方案缓解了这一问题，但是混合同化方法在每个同化时次仍需要一定的集合预报结果作为集合背景场误差协方差的计算样本，如需要120个集合成员的样本，就需要进行120次的模式预报。这对于一些计算条件不是十分充裕的研究和业务单位而言，依然带来不小的计算压力，更影响业务预报效率。x i is the i-th ensemble forecast member, N is the number of ensemble forecast members, Average for ensemble forecasts. It can be seen from formula (2) that when the ensemble forecast error covariance is introduced into the mixture, the calculation of the ensemble members is required, and if the ensemble members are too few, the ensemble forecast error covariance will be dissatisfied with the rank and the variables are not coordinated. Although the mixture The assimilation scheme alleviates this problem, but the hybrid assimilation method still needs a certain amount of ensemble forecast results as the calculation samples of the error covariance of the ensemble background field at each assimilation time. If samples of 120 ensemble members are required, 120 times model forecast. For some research and business units whose computing conditions are not very sufficient, this still brings a lot of computing pressure, and even affects the efficiency of business forecasting.
而目前业务数值预报中，经常要不断地计算和连续保存过去历史预报的相关结果信息，是否能够利用这些历史预报结果作为集合样本，以用于计算集合背景场误差协方差从而用于混合同化，进而提高业务预报的效率，成为本发明所要解决的问题。However, in the current operational numerical forecasting, it is often necessary to continuously calculate and store the relevant result information of the past historical forecast. Whether it is possible to use these historical forecast results as a collection sample to calculate the error covariance of the collection background field for hybrid assimilation? Further improving the efficiency of business forecasting becomes the problem to be solved by the present invention.
发明内容Contents of the invention
本发明所要解决的技术问题是：提供一种基于历史预报结果的准集合-变分的混合资料同化方法，有效引入了各向异性背景场误差协方差，同时又有效降低预报集合带来的计算量。The technical problem to be solved by the present invention is to provide a quasi-ensemble-variational mixed data assimilation method based on historical forecast results, which effectively introduces the error covariance of the anisotropic background field, and at the same time effectively reduces the calculation of the forecast ensemble. quantity.
本发明为解决上述技术问题采用以下技术方案：The present invention adopts the following technical solutions for solving the problems of the technologies described above:
一种基于准集合-变分的混合资料同化方法，包括如下步骤：A method for assimilating mixed data based on quasi-set-variation, comprising the following steps:
步骤1，选取当前预报时刻相邻的历史预报数据，将该数据作为准集合预报样本；Step 1, select the historical forecast data adjacent to the current forecast time, and use this data as a quasi-ensemble forecast sample;
步骤2，对步骤1得到的准集合预报样本，计算同一时刻24小时预报与12小时预报的差异，得到准集合预报误差；Step 2. For the quasi-ensemble forecast sample obtained in step 1, calculate the difference between the 24-hour forecast and the 12-hour forecast at the same time to obtain the quasi-ensemble forecast error;
步骤3，计算步骤2得到的准集合预报误差的均值，将该均值以及准集合预报误差代入公式得到准集合预报误差的无偏估计 Step 3, calculate the mean value of the quasi-ensemble forecast error obtained in step 2, and substitute the mean value and the quasi-ensemble forecast error into the formula Get an unbiased estimate of the quasi-ensemble forecast error
步骤4，将步骤3得到的代入公式将该公式代入集合-变分同化算法中，进行混合同化，优化该算法的目标函数，得到最优分析场；Step 4, the obtained step 3 Into the formula Substituting this formula into the set-variational assimilation algorithm, performing mixed assimilation, optimizing the objective function of the algorithm, and obtaining the optimal analysis field;
其中，为第i次的同一时刻24小时预报与12小时预报的差异，i＝1,…,M，M为准集合预报误差的总数，为准集合预报误差的均值，δx为同化总分析增量，δx1＝x-xb，δx1为三维变分同化的增量，x为分析场，xb为背景场，αi为集合扩展控制变量。in, is the difference between the 24-hour forecast and the 12-hour forecast at the same moment for the ith time, i=1,...,M, M is the total number of quasi-ensemble forecast errors, is the mean value of the quasi-ensemble forecast error, δx is the total analysis increment of assimilation, δx 1 =xx b , δx 1 is the increment of three-dimensional variational assimilation, x is the analysis field, x b is the background field, and α i is the ensemble expansion control variable.
作为本发明的优选方案，所述步骤1的具体过程如下：选取当前预报时刻相邻的过去连续一个月的历史预报数据，每6小时进行的24小时历史预报结果中，提取12小时和24小时的预报数据，将该数据作为准集合预报样本，共计240个。As a preferred solution of the present invention, the specific process of step 1 is as follows: select the historical forecast data of the past continuous month adjacent to the current forecast moment, and extract 12 hours and 24 hours from the 24-hour historical forecast results carried out every 6 hours Forecast data of , which is used as quasi-ensemble forecast samples, a total of 240.
作为本发明的优选方案，所述步骤2的具体过程如下：计算步骤1提取的同一时刻24小时预报与12小时预报的两两差异，得到120个准集合预报误差。As a preferred solution of the present invention, the specific process of step 2 is as follows: Calculate the pairwise difference between the 24-hour forecast and the 12-hour forecast at the same time extracted in step 1 to obtain 120 quasi-ensemble forecast errors.
作为本发明的优选方案，步骤3所述的公式为：其中，分别为24小时预报数据、12小时预报数据。As a preferred version of the present invention, described in step 3 The formula is: in, They are 24-hour forecast data and 12-hour forecast data.
作为本发明的优选方案，步骤3所述M＝120。As a preferred solution of the present invention, M=120 in step 3.
作为本发明的优选方案，步骤4所述准集合-变分同化算法的目标函数为：As a preferred solution of the present invention, the objective function of the quasi-set-variational assimilation algorithm described in step 4 is:
其中，J为目标函数，β1为静态协方差的权重系数，B为静态背景误差协方差矩阵，β2为流依赖协方差的权重系数，α为集合扩展控制变量的向量，A为变量相关矩阵，H为观测算子，R为观测误差协方差矩阵，d＝y-H(xb)为观测增量，其中，y为观测场。Among them, J is the objective function, β1 is the weight coefficient of the static covariance, B is the static background error covariance matrix, β2 is the weight coefficient of the flow-dependent covariance, α is the vector of the set expansion control variables, and A is the variable correlation matrix, H is the observation operator, R is the observation error covariance matrix, d=yH(x b ) is the observation increment, where y is the observation field.
本发明采用以上技术方案与现有技术相比，具有以下技术效果：Compared with the prior art, the present invention adopts the above technical scheme and has the following technical effects:
1、本发明基于准集合-变分的混合资料同化方法，通过历史预报结果中准集合预报误差协方差的引入，给同化系统带来了各向异性、非均质的背景误差协方差信息并且建立和水汽场与其他控制变量的相关关系，使得同化系统能够带来更为合理的同化结果。1. The present invention is based on the quasi-ensemble-variation mixed data assimilation method, through the introduction of the quasi-ensemble forecast error covariance in the historical forecast results, anisotropic and heterogeneous background error covariance information is brought to the assimilation system and Establishing the correlation between the water vapor field and other control variables enables the assimilation system to bring more reasonable assimilation results.
2、本发明基于准集合-变分的混合资料同化方法，准集合预报误差协方差来自于相邻时刻的历史预报数据，通过计算该历史预报数据中，同一时刻24小时预报与12小时预报的差异作为准集合预报误差。该准集合预报误差不是通过集合预报产生，而是通过历史预报结果产生，因此不需要集合预报，计算量与三维变分相当在有效提高预报效果的基础上，还大大节约了计算资源。2. The present invention is based on the quasi-ensemble-variational mixed data assimilation method. The quasi-ensemble forecast error covariance comes from the historical forecast data at adjacent moments. By calculating the historical forecast data, the 24-hour forecast and the 12-hour forecast at the same time The difference is taken as the quasi-ensemble forecast error. The quasi-ensemble forecast error is not generated by the ensemble forecast, but by the historical forecast results, so the ensemble forecast is not needed, and the calculation amount is equivalent to that of the three-dimensional variation. On the basis of effectively improving the forecast effect, it also greatly saves computing resources.
附图说明Description of drawings
图1是本发明基于准集合-变分的混合资料同化方法的操作流程图。Fig. 1 is the operation flowchart of the mixed data assimilation method based on quasi-set-variation in the present invention.
具体实施方式Detailed ways
下面详细描述本发明的实施方式，所述实施方式的示例在附图中示出，其中自始至终相同或类似的标号表示相同或类似的元件或具有相同或类似功能的元件。下面通过参考附图描述的实施方式是示例性的，仅用于解释本发明，而不能解释为对本发明的限制。Embodiments of the present invention are described in detail below, examples of which are shown in the drawings, wherein the same or similar reference numerals denote the same or similar elements or elements having the same or similar functions throughout. The embodiments described below by referring to the figures are exemplary only for explaining the present invention and should not be construed as limiting the present invention.
为了有效引入各向异性背景场误差协方差，同时又有效降低集合预报带来的计算量，建立一种不依赖于集合预报的，既结合三维变分法便于同化多种资料优点，又具有空间各向异性和非均匀性背景误差协方差的同化方案。本发明通过计算历史预报结果误差，得到各向异性、非均质的准集合背景误差协方差，并与三维变分的背景误差协方差相结合，提出了一种准集合-变分混合资料同化方法。In order to effectively introduce the error covariance of the anisotropic background field and effectively reduce the calculation amount brought by ensemble forecasting, a method that does not depend on ensemble forecasting is established. Assimilation scheme for anisotropic and non-uniform background error covariance. The present invention obtains the anisotropic and heterogeneous quasi-ensemble background error covariance by calculating the error of historical forecast results, and combines it with the background error covariance of the three-dimensional variation to propose a quasi-set-variational mixed data assimilation method.
为降低预报集合带来的计算量，同时又能够引入各向异性和非均质的背景场误差协方差，将历史预报样本中同一时刻不同时效的模式预报场的差异，作为一个准集合预报误差，准集合预报误差的预报误差集合是由连续一段时间同一个时刻不同预报时效的预报误差构成。In order to reduce the amount of calculation brought by the forecast ensemble, and at the same time introduce anisotropic and heterogeneous background field error covariance, the difference of the model forecast field at the same time and different timeliness in the historical forecast sample is taken as a quasi-ensemble forecast error , the forecast error set of the quasi-ensemble forecast error is composed of forecast errors with different forecast timeliness at the same moment for a continuous period of time.
资料同化的目的就是寻找一个最优的分析场使得目标函数最小，集合-变分混合同化方法的目标函数可表示为：The purpose of data assimilation is to find an optimal analysis field to minimize the objective function. The objective function of the ensemble-variation hybrid assimilation method can be expressed as:
其中，各字母含义同上。Wherein, the meaning of each letter is the same as above.
对于常规混合同化，式(1)中为集合预报误差的无偏估计：For conventional mixed assimilation, in formula (1) is an unbiased estimate of the ensemble forecast error:
其中，各字母含义同上。Wherein, the meaning of each letter is the same as above.
为降低预报集合带来的计算量，同时又能够引入各向异性和非均质的背景场误差协方差，本发明将历史预报样本中同一时刻不同时效的模式预报场的差异，作为一个准集合预报误差协方差，即将视为模式准集合预报误差集合的无偏估计：In order to reduce the amount of calculation brought by the forecast set, and at the same time introduce anisotropic and heterogeneous background field error covariance, the present invention regards the difference of the model forecast field with different timeliness at the same time in the historical forecast samples as a quasi-set Forecast error covariance, which is Treated as an unbiased estimate of the ensemble of model quasi-ensemble forecast errors:
其中，为第i时次的同一时刻不同时效的模式预报场的差异：in, is the difference of the model forecast field with different timeliness at the same time at the i-th time:
T1、T2为预报时效，这里为历史预报的时间平均(目的是消除时间平均偏差)，M为历史预报误差总数，M可以等于N。从上述推导可以明显看出，本发明的方法既避免了集合预报带来的较大计算量，又使得同化系统具备了各向异性和非均匀性的背景场误差协方差信息。T1 and T2 are forecast timeliness, here is the time average of historical forecast (the purpose is to eliminate the time average deviation), M is the total number of historical forecast errors, and M can be equal to N. It can be clearly seen from the above derivation that the method of the present invention not only avoids the large amount of calculation brought by the ensemble forecast, but also enables the assimilation system to have anisotropic and non-uniform background field error covariance information.
如图1所示，历史预报数据一般选取相邻时刻的过去连续一个月的时间，当然也可以是15天或45天等，但从统计样本的充足性考虑，一个月的数据更为合理。以2011年7月16日12时的集合-变分混合同化为例，本发明基于准集合-变分的混合资料同化方法包括如下步骤：As shown in Figure 1, the historical forecast data generally selects the past consecutive month of adjacent moments, of course, it can also be 15 days or 45 days, etc., but considering the adequacy of statistical samples, the data of one month is more reasonable. Taking the set-variation mixed assimilation at 12 o'clock on July 16, 2011 as an example, the present invention based on the quasi-set-variation mixed data assimilation method includes the following steps:
步骤1，先提取相邻一个月的连续历史预报结果，以获得历史预报样本。即：从2011年6月15日00时到2011年7月16日00时，在每6小时进行一次的历史预报结果中，提取12小时和24小时的预报结果，将该结果作为准集合预报样本，即从2011年6月15日00时起报的预报结果中，提取12时和24时(即次日00时)的预报结果，2011年6月15日06时起报的预报结果中，提取18时和次日06时的预报结果，依次类推。Step 1, first extract the continuous historical forecast results of adjacent months to obtain historical forecast samples. That is: from 00:00 on June 15, 2011 to 00:00 on July 16, 2011, extract the 12-hour and 24-hour forecast results from the historical forecast results every 6 hours, and use the results as quasi-ensemble forecasts Sample, that is, from the forecast results reported at 00:00 on June 15, 2011, the forecast results at 12:00 and 24:00 (that is, 00:00 the next day) are extracted, and the forecast results reported at 06:00 on June 15, 2011 , extract the forecast results at 18:00 and 06:00 the next day, and so on.
步骤2，在这连续一个月内获得的准集合预报样本中，计算同一时刻不同时效的模式预报场的差异作为“准集合预报误差”。如，从2011年6月15日00时到2011年7月16日00时，每6小时一次的历史预报结果中，选取24小时预报与12小时预报的差异作为预报误差(即2011年6月15日00时起报24小时，可以得到2011年6月16日00时的结果，2011年6月15日12起报12小时，也可以得到2011年6月16日00时的结果，将上述两个预报结果的差异作为预报误差)，每6小时一个得到误差场，那么这一个月就可以得到120个预报误差样本。该历史预报误差集合不是通过集合预报产生，而是通过一个月的连续历史预报结果产生，从而将该预报误差称为“准集合预报误差”。Step 2: In the quasi-ensemble forecast samples obtained in this continuous month, calculate the difference between the model forecast fields with different timeliness at the same time as the “quasi-ensemble forecast error”. For example, from 00:00 on June 15, 2011 to 00:00 on July 16, 2011, in the historical forecast results every 6 hours, the difference between the 24-hour forecast and the 12-hour forecast is selected as the forecast error (that is, in June 2011 Report for 24 hours from 00:00 on June 15, 2011, and you can get the result at 00:00 on June 16, 2011. Report for 12 hours from 12 on June 15, 2011, and you can also get the result at 00:00 on June 16, 2011. The difference between the two forecast results is regarded as the forecast error), and the error field is obtained every 6 hours, so 120 forecast error samples can be obtained in this month. This set of historical forecast errors is not generated by ensemble forecasts, but by one-month continuous historical forecast results, so this forecast error is called "quasi-ensemble forecast error".
步骤3，利用步骤2获得的预报误差集合，计算预报误差的均值，并将预报误差集合和均值代入公式(3)进行计算，获得2011年7月16日12时的准预报误差集合的无偏估计。Step 3, use the forecast error set obtained in step 2 to calculate the mean value of the forecast error, and substitute the forecast error set and the mean value into the formula (3) for calculation, and obtain the unbiased estimate.
步骤4，将步骤3获得的“准集合预报误差”，代入公式引入混合同化系统，即公式(1)，与静态协方差B相结合进行正常的2011年7月16日12时的混合同化。Step 4. Substitute the "quasi-ensemble forecast error" obtained in step 3 into the formula Introduce the mixed assimilation system, that is, formula (1), and combine it with the static covariance B to carry out the normal mixed assimilation at 12 o'clock on July 16, 2011.
步骤5，在混合同化时，基于2011年7月16日12时的背景场，同化2011年7月16日12时的观测资料获得该时刻的分析场，在分析场基础上进行数值天气预报获得下一时刻预报场。Step 5. During the hybrid assimilation, based on the background field at 12:00 on July 16, 2011, assimilate the observation data at 12:00 on July 16, 2011 to obtain the analysis field at that time, and perform numerical weather prediction on the basis of the analysis field to obtain Forecast for the next moment.
以上实施例仅为说明本发明的技术思想，不能以此限定本发明的保护范围，凡是按照本发明提出的技术思想，在技术方案基础上所做的任何改动，均落入本发明保护范围之内。The above embodiments are only to illustrate the technical ideas of the present invention, and can not limit the protection scope of the present invention with this. All technical ideas proposed in accordance with the present invention, any changes made on the basis of technical solutions, all fall within the protection scope of the present invention. Inside.
Claims (6)
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201510353774.7A CN104992057A (en) | 2015-06-25 | 2015-06-25 | Quasi-ensemble-variation based mixed data assimilation method |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201510353774.7A CN104992057A (en) | 2015-06-25 | 2015-06-25 | Quasi-ensemble-variation based mixed data assimilation method |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN104992057A true CN104992057A (en) | 2015-10-21 |
Family
ID=54303871
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201510353774.7A Pending CN104992057A (en) | 2015-06-25 | 2015-06-25 | Quasi-ensemble-variation based mixed data assimilation method |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN104992057A (en) |
Cited By (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN105447593A (en) * | 2015-11-16 | 2016-03-30 | 南京信息工程大学 | Rapid updating mixing assimilation method based on time lag set |
| CN109063083A (en) * | 2016-07-20 | 2018-12-21 | 中国水利水电科学研究院 | A kind of multi-source weather information data assimilation method |
| CN109212631A (en) * | 2018-09-19 | 2019-01-15 | 中国人民解放军国防科技大学 | A 3D Variational Assimilation Method for Satellite Observation Data Considering Channel Correlation |
| CN110020462A (en) * | 2019-03-07 | 2019-07-16 | 江苏无线电厂有限公司 | The method that a kind of pair of meteorological data carries out fusion treatment and generate numerical weather forecast |
| CN110110922A (en) * | 2019-04-30 | 2019-08-09 | 南京信息工程大学 | A kind of adaptive partition assimilation method based on rain belt sorting technique |
| CN111783361A (en) * | 2020-07-07 | 2020-10-16 | 中国人民解放军国防科技大学 | Hybrid data assimilation method for numerical weather forecasting based on triple multilayer perceptron |
| CN114070262A (en) * | 2021-10-26 | 2022-02-18 | 南京大学 | An Integrated Mixed Ensemble Kalman Filter Weather Forecast Assimilation Method and Device with Additional Disturbance |
| CN116975523A (en) * | 2023-09-22 | 2023-10-31 | 南京气象科技创新研究院 | Data assimilation background error covariance characteristic statistical method for strong convection weather typing |
Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JP2010060443A (en) * | 2008-09-04 | 2010-03-18 | Japan Weather Association | Weather forecast device, method, and program |
| CN101814117A (en) * | 2010-04-14 | 2010-08-25 | 北京师范大学 | Multi-source environment ecological information data assimilation method |
| CN102034027A (en) * | 2010-12-16 | 2011-04-27 | 南京大学 | Method for assimilating remote sensing data of soil humidity in watershed scale |
| CN102323987A (en) * | 2011-09-09 | 2012-01-18 | 北京农业信息技术研究中心 | A method for assimilating crop leaf area index |
| CN102737155A (en) * | 2011-04-12 | 2012-10-17 | 中国科学院寒区旱区环境与工程研究所 | Bayesian fitering-based general data assimilation method |
| CN104036135A (en) * | 2014-06-06 | 2014-09-10 | 南京大学 | Typhoon dynamic balance constrained variational assimilation method based on WRF (Weather Research and Forecasting) mode |
-
2015
- 2015-06-25 CN CN201510353774.7A patent/CN104992057A/en active Pending
Patent Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JP2010060443A (en) * | 2008-09-04 | 2010-03-18 | Japan Weather Association | Weather forecast device, method, and program |
| CN101814117A (en) * | 2010-04-14 | 2010-08-25 | 北京师范大学 | Multi-source environment ecological information data assimilation method |
| CN102034027A (en) * | 2010-12-16 | 2011-04-27 | 南京大学 | Method for assimilating remote sensing data of soil humidity in watershed scale |
| CN102737155A (en) * | 2011-04-12 | 2012-10-17 | 中国科学院寒区旱区环境与工程研究所 | Bayesian fitering-based general data assimilation method |
| CN102323987A (en) * | 2011-09-09 | 2012-01-18 | 北京农业信息技术研究中心 | A method for assimilating crop leaf area index |
| CN104036135A (en) * | 2014-06-06 | 2014-09-10 | 南京大学 | Typhoon dynamic balance constrained variational assimilation method based on WRF (Weather Research and Forecasting) mode |
Non-Patent Citations (2)
| Title |
|---|
| 王铁 等: "混合数据资料同化方法在WRF模式中的应用", 《第27届中国气象学会年会灾害天气研究与预报分会场论文集》 * |
| 陈耀登 等: "一种控制变量循环扰动和多参数化方案的混合同化方法", 《大气科学学报》 * |
Cited By (15)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN105447593A (en) * | 2015-11-16 | 2016-03-30 | 南京信息工程大学 | Rapid updating mixing assimilation method based on time lag set |
| CN109063083A (en) * | 2016-07-20 | 2018-12-21 | 中国水利水电科学研究院 | A kind of multi-source weather information data assimilation method |
| CN109063083B (en) * | 2016-07-20 | 2021-05-07 | 中国水利水电科学研究院 | A Data Assimilation Method of Multi-source Meteorological Information |
| CN109212631B (en) * | 2018-09-19 | 2020-12-01 | 中国人民解放军国防科技大学 | A 3D Variational Assimilation Method for Satellite Observation Data Considering Channel Correlation |
| CN109212631A (en) * | 2018-09-19 | 2019-01-15 | 中国人民解放军国防科技大学 | A 3D Variational Assimilation Method for Satellite Observation Data Considering Channel Correlation |
| CN110020462A (en) * | 2019-03-07 | 2019-07-16 | 江苏无线电厂有限公司 | The method that a kind of pair of meteorological data carries out fusion treatment and generate numerical weather forecast |
| CN110020462B (en) * | 2019-03-07 | 2023-04-07 | 江苏无线电厂有限公司 | Method for fusing meteorological data and generating numerical weather forecast |
| CN110110922A (en) * | 2019-04-30 | 2019-08-09 | 南京信息工程大学 | A kind of adaptive partition assimilation method based on rain belt sorting technique |
| CN110110922B (en) * | 2019-04-30 | 2023-06-06 | 南京信息工程大学 | An Adaptive Partition Assimilation Method Based on Rain Region Classification Technology |
| CN111783361A (en) * | 2020-07-07 | 2020-10-16 | 中国人民解放军国防科技大学 | Hybrid data assimilation method for numerical weather forecasting based on triple multilayer perceptron |
| CN111783361B (en) * | 2020-07-07 | 2021-03-12 | 中国人民解放军国防科技大学 | Hybrid data assimilation method for numerical weather forecasting based on triple multilayer perceptron |
| CN114070262A (en) * | 2021-10-26 | 2022-02-18 | 南京大学 | An Integrated Mixed Ensemble Kalman Filter Weather Forecast Assimilation Method and Device with Additional Disturbance |
| CN114070262B (en) * | 2021-10-26 | 2022-06-21 | 南京大学 | An Integrated Mixed Ensemble Kalman Filter Weather Forecast Assimilation Method and Device with Additional Disturbance |
| CN116975523A (en) * | 2023-09-22 | 2023-10-31 | 南京气象科技创新研究院 | Data assimilation background error covariance characteristic statistical method for strong convection weather typing |
| CN116975523B (en) * | 2023-09-22 | 2023-12-12 | 南京气象科技创新研究院 | Data assimilation background error covariance characteristic statistical method for strong convection weather typing |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN104992057A (en) | Quasi-ensemble-variation based mixed data assimilation method | |
| Abbasian et al. | Performance of the general circulation models in simulating temperature and precipitation over Iran | |
| Al Kajbaf et al. | Application of surrogate models in estimation of storm surge: A comparative assessment | |
| Huang et al. | Monthly streamflow prediction using modified EMD-based support vector machine | |
| Wang et al. | Analysis of spatio-temporal evolution of droughts in Luanhe River Basin using different drought indices | |
| Fant et al. | Projections of water stress based on an ensemble of socioeconomic growth and climate change scenarios: a case study in Asia | |
| Shi et al. | Physically based simulation of the streamflow decrease caused by sediment‐trapping dams in the middle Yellow River | |
| Chen et al. | A new method for identification of flood seasons using directional statistics | |
| Zhang et al. | Combining postprocessed ensemble weather forecasts and multiple hydrological models for ensemble streamflow predictions | |
| CN108171007A (en) | One kind is based on the modified Medium-and Long-Term Runoff Forecasting method of numerical value DATA PROCESSING IN ENSEMBLE PREDICTION SYSTEM extreme value | |
| Verkade et al. | Estimating predictive hydrological uncertainty by dressing deterministic and ensemble forecasts; a comparison, with application to Meuse and Rhine | |
| WO2018214364A1 (en) | Precipitation forecast accuracy assessment method | |
| Guo et al. | Prediction of daily crop reference evapotranspiration (ET0) values through a least-squares support vector machine model | |
| CN105447593A (en) | Rapid updating mixing assimilation method based on time lag set | |
| CN104040378B (en) | Weather prognosis device and Predictive meteorological methods | |
| Li et al. | Real-time flood simulations using CA model driven by dynamic observation data | |
| CN115983511B (en) | Precipitation prediction method and system based on improved statistical downscaling method | |
| Chen et al. | Urban inundation rapid prediction method based on multi-machine learning algorithm and rain pattern analysis | |
| Guo et al. | Synergistic evolution and attribution analysis of water-sediment in the middle and lower reaches of the Yangtze River | |
| CN107944466B (en) | Rainfall deviation correction method based on segmentation idea | |
| CN114239274A (en) | Method for calculating root layer soil water by multi-source remote sensing data driven index filtering model | |
| CN118036889B (en) | Multi-time-space scale evolution feature analysis method and system for different types of drought | |
| CN115454984A (en) | Satellite precipitation data correction method with self-adjusting window | |
| Zhang et al. | Sustainable effect of coal mining on hydrological drought risk: a case study on Kuye river basin, China | |
| Abaza et al. | Hydrological evaluation of the Canadian meteorological ensemble reforecast product |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| WD01 | Invention patent application deemed withdrawn after publication |
Application publication date: 20151021 |
|
| WD01 | Invention patent application deemed withdrawn after publication |