The present invention relates to the technical field of precipitation forecasting, in particular to a method for associating a precipitation forecast capability with a teleconnection effect based on the coefficient of determination.
Accurate seasonal precipitation forecast has important value and broad application prospects in the fields of disaster prevention and mitigation of natural disasters such as floods and droughts, and in the fields of water resources planning and management, etc. El Niño-Southern Oscillation (ENSO) is an important driving factor of global seasonal precipitation anomalies. In some studies, climate modes characterizing ENSO events is used as predictors and seasonal precipitation is predicted by means of teleconnection effects of climate modes (such as Niño3.4, Niño3 and Niño4). In addition, meteorological and climate centers in many countries and regions in the world have begun to research and develop their own global climate models (GCMs). These models characterize various key physical processes related to climate, and their forecast results have clear physical meanings.
Although climate modes and GCM precipitation forecast can provide information for seasonal precipitation, when it comes to practical application of these information, it is difficult to judge whether GCM seasonal precipitation with a certain physical meaning includes information of key teleconnection effects; and it is difficult to judge that in different regions of the world to what extent information provided respectively by the climate modes and the GCM seasonal precipitation is redundant, and that how much information is respectively provided by the climate modes and the GCM seasonal precipitation. In order to answer these questions, it is necessary to provide a simple and practical method to judge the overlap and difference between the information of the climate modes and the GCM seasonal precipitation, so as to provide a certain reference for the use of forecast information in practical services.
In order to solve the problem in judging the overlap and difference between information of global climate model based seasonal precipitation forecast and key teleconnection effects, the present invention provides a method for associating a precipitation forecast capability with a teleconnection effect based on coefficients of determination.
In order to solve the above technical problem, the present invention adopts the following technical solution:
A method for associating a precipitation forecast capability with a teleconnection effect based on coefficients of determination, including following steps:
Compared with the prior art, the technical solution of the present invention has the following beneficial effects: the present invention combines the set operation with the coefficients of determination in linear regression, and simply and effectively distinguishes overlapped and different components in observed precipitation information provided by the precipitation forecast and the climate modes, thereby providing a reference for the use of a service of the precipitation forecast.
The accompanying drawings are merely used for exemplary description, and should not be construed as a limitation to the present patent.
For those skilled in the art, it is understandable that some well-known structures in the accompanying drawings and their descriptions may be omitted.
The technical solution of the present invention is further described below with reference to the accompanying drawings and the embodiments.
The present embodiment provides a method for associating a precipitation forecast capability with a teleconnection effect based on coefficients of determination.
The method for associating the precipitation forecast capability with the teleconnection effect based on the coefficients of determination provided by the present embodiment includes the following steps:
A climate index contained in the climate index sample sequence includes niño3.4, niño3 and/or niño4. In the present embodiment, the niño3.4 is selected and used as the climate index.
In this step, the regression equation of the observed precipitation ok and the forecast precipitation fk, the regression equation of the observed precipitation ok and the climate index niño3.4k, and the regression equation of the observed precipitation ok and the union set (f∪niño3.4) of the forecast precipitation and the climate index are respectively established, and the coefficients of determination determined by the above three regression equations are further calculated.
A calculation formula for a coefficient of determination R2(o˜f) determined by the regression equation of the observed precipitation ok and the forecast precipitation fk is as follows:
A calculation formula for a coefficient of determination R2(o˜niño3.4) determined by the regression equation of the observed precipitation ok and the climate index niño3.4k is as follows:
A calculation formula for a coefficient of determination R2(o˜f∪niño3.4) determined by the regression equation of the observed precipitation ok and the union set (f∪niño3.4) of the forecast precipitation and the climate index is as follows:
An expression formula for the variance explained by the forecast precipitation alone is as follows:
An expression formula for the variance explained by the climate index alone is as follows:
An expression formula for the variance repeatedly explained by the forecast precipitation and the climate index alone is as follows:
In the present embodiment, the step of processing the variances by means of bootstrapping includes: the historical forecast precipitation data and the climate index sample sequence are disrupted, the steps S2 to S3 are repeated to obtain the corresponding three variances, and until the preset number of iterations is reached, the reference distribution of the three variances is obtained.
In the present embodiment, the set number of iterations is 1,000.
In the present step, the original sample data is compared with the reference distribution of the three variances by means of a one-sided test. The step of comparing the original sample data with the reference distribution of the three variances includes: a significance level is selected, where 0.1, 0.05, and 0.01 may be selected to be the significance levels in general, and reference distribution thresholds corresponding to the significance levels are respectively 90th, 95th, and 99th percentiles of the reference distribution; when the significance level is set to be 0.1, if a value of the original sample data is greater than the 90th percentile of the reference distribution of its corresponding variances, it identifies that the value of the original sample data is significant, otherwise it identifies that the value of the original sample data is non-significant. And then a significance result is output as the association result of the precipitation forecast capability and the teleconnection effect.
In a specific implementation process, there are 8 significant results, which are specifically as shown in Table 1 below.
1 represents that the value of the original sample data is identified to be significant at the corresponding coefficient of determination, and 0 represents that the value of the original sample data is identified to be non-significant at the corresponding coefficient of determination.
In the present embodiment, the set operation is combined with the coefficients of determination in linear regression, and are further combined with the bootstrapping and the one-sided test, so as to simply and effectively distinguish overlapped and different components in observed precipitation information provided by the precipitation forecast and the climate mode, thereby providing a reference for the use of a service of the precipitation forecast.
In the present embodiment, a test is performed based on the method for associating a precipitation forecast capability with a teleconnection effect provided by the embodiment 1.
1982-2010 global seasonal grid precipitation data of the United States Climate Prediction Center (CPC) is used as observed data, a climate forecast system Version 2 (CFSv2) of the United States National Centers for Environmental Prediction (NCEP) is used as forecast precipitation data, and an index Niño3.4 is used to represent El Niño-Southern Oscillation (ENSO). CFSv2 forecast precipitation adopts seasonal forecast precipitation with a forecast period of 0 month. The winter December-January-February (DJF) is taken as an example. A spatial resolution of both observed precipitation and forecast precipitation is 1°×1°.
For precipitation of each grid in
In order to quantify information respectively provided by the forecast precipitation and the Niño3.4 and overlapped information of the forecast precipitation and the Niño3.4, further, overlapped R2 of the forecast precipitation and the Niño3.4 and the respectively independent R2 of the forecast precipitation and the Niño3.4 are obtained by means of set operation.
Further, in the present embodiment, three groups of variances R2(o˜f/niño3.4), R2(o˜niño3.4/f), and R2(o˜f∩niño3.4) determined according to coefficients of determination are respectively subjected to a significance test. There may be eight different combinations for three groups of significance results, specifically as shown in Table 1.
Further, as shown in
The above experimental results show that the method for associating a precipitation forecast capability with a teleconnection effect based on coefficients of determination provided by the present invention can effectively quantify the difference and overlap between the information provided by the forecast precipitation and the index Niño3.4, and can intuitively show different association relationship classification conditions, thereby providing a reference for the use of a service of the forecast.
Apparently, the above-mentioned embodiments of the present invention are only examples for clearly illustrating the present invention, and are not intended to limit the implementation modes of the present invention. Those of ordinary skill in the art may also make other changes or modifications in different forms on the basis of the above description. All implementation modes do not need to and cannot be exhausted here. Any modifications, equivalent substitutions, and improvements made within the spirit and principle of the present invention should be included within the scope of protection of the claims of the present invention.
Filing Document | Filing Date | Country | Kind |
---|---|---|---|
PCT/CN2021/123451 | 10/13/2021 | WO |