A Method and System for Correcting Precipitation in Subseasonal Models Based on Climate Zoning and Sea Temperature Background Constraints

Abstract

This invention relates to the field of meteorological services and provides a method and system for correcting subseasonal model precipitation based on climate zoning and sea surface temperature (SST) background constraints. The method includes: climate zoning, adding SST background constraints, reanalysis data modeling, model historical return data modeling, real-time prediction model matching, and prediction result correction. The system includes: a data preprocessing unit, a climate zoning unit, an SST background constraint unit, a reanalysis data modeling unit, a model historical return data modeling unit, and a real-time model matching and prediction result correction unit. This invention considers the importance of regional differences in precipitation and SST background constraints, solving the problems of significant regional differences in daily summer precipitation in my country, the tendency for overfitting in single-point modeling, and the lack of physical constraints that make correction methods unsuitable. It effectively utilizes physically meaningful predictability sources to modify the cumulative probability density correction method based on percentile mapping, thereby improving my country's summer subseasonal precipitation prediction capabilities.

Description

Technical Field

[0001] This invention relates to the field of meteorological services, and in particular to a method and system for correcting subseasonal model precipitation based on climate zoning and sea surface temperature background constraints. It takes into account both climate zoning factors and sea surface temperature background factors to correct the subseasonal model precipitation prediction results, thereby significantly improving the accuracy of the prediction results. Background Technology

[0002] With the frequent occurrence of extreme weather events, climate forecasts on a sub-seasonal timescale of two weeks to two months can provide early warnings of extreme events and offer strong scientific and technological support for disaster prevention and mitigation. However, the current level of sub-seasonal forecasting is still significantly lower than that of short-term weather forecasts and medium- to long-term climate predictions. In the field of sub-seasonal precipitation forecasting, both domestic and international methods mainly employ dynamical and statistical forecasting, but both methods have limitations in their predictive capabilities. Specifically, mainstream sub-seasonal to seasonal (S2S) dynamical models, such as those of the European Centre for Medium-Range Weather Forecasts (ECMWF), typically maintain their precipitation forecasting skill for only about two weeks (Domeisen et al., 2022), and the skill declines rapidly with the extension of the forecast period, particularly regarding the forecasting ability for heavy precipitation events. Furthermore, while statistical models (e.g., spatiotemporal projection STPM, Zhu and Li, 2017) or dynamical-statistical combined forecasting models (Wu et al., 2022) have some predictive ability for sub-seasonal precipitation trends, they exhibit significant biases in predicting the central location and intensity of precipitation events. Overall, current sub-seasonal climate forecasting technologies are insufficient to meet the significant service needs of meteorological disaster prevention and mitigation.

[0003] Model error correction is a key technique that combines historical long-term observations with model return data, correcting the direct model output through the statistical characteristics of historical observations to improve forecasting accuracy. In particular, the percentile mapping correction method (QM) based on cumulative probability density (CDF) primarily corrects the cumulative probability density curve of the element being corrected, showing significant correction effects on mean and extreme values. For example, Zhang Daquan and Chen Lijuan (2016) corrected the monthly mean temperature of the DERF2.0 model, effectively reducing the root mean square error of the model forecast, and also improving the temperature anomaly distribution forecast to varying degrees. However, the existing CDF-QM method is not suitable for correcting the sub-seasonal model forecasts of daily precipitation in my country. On the one hand, the probability distribution of precipitation exhibits skewed distribution characteristics, which differs significantly from the normal distribution characteristics of temperature. At the same time, precipitation is significantly affected by topography, and its spatial distribution characteristics are relatively complex. Correcting each grid point individually results in a small sample size, which easily leads to overfitting. On the other hand, in the absence of physical background constraints, the CDF-QM method can only correct systematic biases and cannot reflect the modulation effect of external forcing signals such as sea surface temperature on the precipitation process.

[0004] Solving the aforementioned technical problems requires appropriately increasing the sample size while considering reasonable physical background constraints. When increasing the sample size required for modeling, a sufficient quantity must be achieved to avoid overfitting, while also ensuring consistency among samples to prevent the model from being inapplicable to the grid points requiring correction. When considering physical background constraints, the physical signal must possess characteristics such as stability, slow variation, and significant influence. Therefore, addressing existing technical problems and improving the predictive ability of sub-seasonal models for summer precipitation in my country requires substantial meteorological knowledge and presents a certain level of technical difficulty. Solving this technical problem can effectively leverage information from existing sub-seasonal dynamic models, significantly improving the predictive ability for summer sub-seasonal precipitation in my country, and providing crucial scientific and technological support for disaster prevention and mitigation and water resource management. Summary of the Invention

[0005] The purpose of this invention is to overcome the shortcomings of existing technologies. Addressing the problems of significant regional differences in daily precipitation in summer in my country, the tendency for overfitting in single-point modeling, and the inapplicability of the CDF-QM correction method due to the lack of physical constraints, this invention provides a sub-seasonal model precipitation correction method and system based on climate zoning and sea surface temperature background constraints. This method corrects errors in the sub-seasonal model prediction results of daily precipitation in summer in my country, thereby improving the ability to predict sub-seasonal precipitation in summer in my country.

[0006] Glossary of terms in this article:

[0007] Further analysis of precipitation data: a spatiotemporally continuous precipitation dataset generated by integrating multiple observational data and numerical model simulation results.

[0008] Historical data from the model: The past time point 1 is the starting date, and the precipitation data for the past time point 2 is predicted using the next seasonal model;

[0009] Precipitation forecast results: Precipitation data for future time points are predicted using the sub-seasonal model.

[0010] Prediction for a forecast date on a given starting date: Using a given starting date x as the starting point, predict the precipitation amount for a forecast date y.

[0011] The present invention adopts the following technical solution:

[0012] On the one hand, this invention provides a method for correcting subseasonal model precipitation based on climate zoning and sea surface temperature background constraints, comprising:

[0013] S1. Climate Zoning: The area to be corrected is gridded at a certain resolution; based on the daily characteristics of summer precipitation, the area to be corrected is clustered into several climate sub-regions using reanalysis precipitation data.

[0014] S2. Add sea surface temperature background constraints: Using the sea surface temperature background model, add sea surface temperature background constraints to all the climate sub-regions mentioned in step S1.

[0015] S3. Reanalysis Data Modeling: For each climate sub-region with sea surface temperature background constraints, a sliding window CDF model is performed based on the reanalysis precipitation data of a certain date within a historical period to obtain the first CDF curve of the climate sub-region under the constraints of the sea surface temperature background on that date.

[0016] S4. Modeling historical return data: For each climate sub-region with sea surface temperature background constraints, based on the sub-seasonal model, a sliding window CDF model is performed on all historical return data of the model on a forecast date of a certain reporting date to obtain the second CDF curve of the climate sub-region under the constraints of the reporting date, the forecast date, and the sea surface temperature background.

[0017] S5. Real-time prediction model matching: Determine the sea surface temperature background constraint of the current year, and based on the sub-seasonal model, project the grid-point precipitation prediction results of a certain climate sub-region, a certain reporting date, and a certain forecast date onto the second CDF curve of the same climate sub-region, the same sea surface temperature background constraint, the same reporting date, and the same forecast date as in step S4, to obtain the probability density percentile position corresponding to the precipitation prediction results.

[0018] S6. Prediction result correction: The probability density percentile position obtained in step S5 is mapped to the first CDF curve with the same climate sub-region, the same sea surface temperature background constraint, and the same date as the forecast date obtained in step S3, to obtain the reanalysis precipitation value corresponding to the probability density percentile position. This reanalysis precipitation value is used as the corrected prediction result for the forecast date of the certain reporting date.

[0019] In addition to any of the possible implementations described above, a further implementation is provided, wherein the method further includes:

[0020] S7. Repeat steps S5 and S6 to perform real-time prediction model matching and prediction result correction for each prediction day of a given reporting date, and obtain the corrected prediction results for all prediction days of a given reporting date.

[0021] In addition to any of the possible implementations described above, a further implementation is provided in which, in step S1, K-means clustering is used to classify the region to be corrected; all grid points are divided into K groups, K grid points are randomly selected as initial cluster centers, and then the distance (mean squared error) between each grid point and the precipitation time series of other grid points is calculated, and each grid point is assigned to the nearest cluster center; the cluster centers and the grid points assigned to them represent a cluster; each time a grid point is assigned, the cluster center is recalculated based on all existing grid points in the cluster; this process is repeated until no grid points are reassigned to different clusters, no cluster centers change, and the sum of squared errors is minimized within the same cluster, thus completing the classification of the region to be corrected; each cluster forms a climate sub-region.

[0022] In addition to any of the possible implementations described above, another implementation is provided in which, in step S2, the area to be corrected is the entire region of China, the sea surface temperature background model adopts the Indian Ocean Basin Consistent Model (IOBM), and the years are divided into normal years, positive anomaly years, and negative anomaly years according to the IOBM index; years with an IOBM index greater than 1.5 times the standard deviation are positive anomaly years, years with an IOBM index less than -1.5 times the standard deviation are negative anomaly years, and years with an IOBM index within ±1.5 times the standard deviation are normal years.

[0023] In addition to any of the possible implementations described above, another implementation is provided in which the IOBM index is defined as the anomaly value of the mean sea surface temperature in the 20°S-20°N and 40°E-110°E region after removing the linear trend.

[0024] The standardized calculation method for the IOBM index is as follows:

[0025] here

[0026] Where, xi Represents the IOBM exponent, y i This represents the standardized IOBM index. Let n be the mean of the IOBM index, and n be the number of IOBM indices.

[0027] In addition to any of the possible implementations described above, another implementation is provided in which the method for obtaining the first CDF curve in step S3 is as follows:

[0028] S31. Set the spacing between precipitation groups, wherein the spacing between precipitation groups is gradually set, and the spacing between small precipitation groups is smaller than the spacing between large precipitation groups.

[0029] S32. For the daily precipitation reanalysis values ​​of all grid points in each climate sub-region of the country from May 1 to September 30 according to the clustering results of step S1 within the sliding window of positive, negative and normal years of IOBM index, calculate the cumulative probability density value of precipitation in each precipitation interval according to the set gradual group interval, and obtain the cumulative probability density curve of the entire 0-100 mm range, i.e. the first CDF curve.

[0030] In addition to any of the possible implementations described above, another implementation is provided in which the method for obtaining the second CDF curve in step S4 is as follows:

[0031] S41. Set the interval between precipitation groups, which is the same as in step S31;

[0032] S42. For the daily forecasts from May 1 to September 30, predicting the next 0-60 days (taking CMA-CPSv3 as an example, the forecast lead time varies for different models), calculate the cumulative probability density value of the model's historical precipitation in each precipitation interval within the entire historical return period (taking CMA-CPSv3 as an example, its historical return period is the past 15 years with the start year as a reference) according to the positive anomaly year, negative anomaly year, and normal year of the IOBM index, based on the pre-set gradual group interval. This yields the cumulative probability density curve in the range of 0-100 mm, i.e., the second CDF curve.

[0033] In addition to any of the possible implementations described above, another implementation is provided in which the spacing between the precipitation groups is set as follows: 0.1 mm in the range of 0-1 mm, 0.2 mm in the range of 1-10 mm, 0.5 mm in the range of 10-25 mm, 1.0 mm in the range of 25-50 mm, and 2.0 mm in the range of 50-100 mm.

[0034] In addition to any of the possible implementations described above, another implementation is provided in which, in step S5, all forecast days (taking CMA-CPSv3 as an example, the result of forecasting the next 0-60 days in one forecast) and all grid points are judged and calculated one by one for each reporting start date.

[0035] On the other hand, the present invention also provides a subseasonal model precipitation correction system based on climate zoning and sea surface temperature background constraints, comprising:

[0036] The data preprocessing unit is used for standardizing precipitation data and gridding the area to be corrected. The standardization of precipitation data specifically includes: processing the cumulative precipitation reported by the sub-seasonal model and the forecast cumulative precipitation into daily average precipitation; interpolating the reanalysis precipitation data, the daily average precipitation reported by the sub-seasonal model, and the forecast daily average precipitation to grids with the same spatial resolution.

[0037] The climate zoning unit, based on the daily characteristics of summer precipitation, uses the reanalysis precipitation data processed by the data preprocessing unit to cluster and divide the area to be corrected into several climate sub-regions.

[0038] The sea surface temperature background constraint unit uses the sea surface temperature background model to add sea surface temperature background constraints to all the climate sub-regions.

[0039] The data modeling unit then performs sliding window CDF modeling on the reanalysis precipitation data of each sub-region with sea surface temperature background constraints on a certain date within a historical period, to obtain the first CDF curve of the climate sub-region under the constraints of sea surface temperature background on that date.

[0040] The modeling unit for historical return data of the model performs sliding window CDF modeling on all historical return data of the model on a forecast date for a certain forecast date based on the sub-seasonal model for each climate sub-region with sea surface temperature background constraints, and obtains the second CDF curve of the climate sub-region under the constraints of the forecast date, the forecast date, and the sea surface temperature background.

[0041] The real-time model matching and prediction result correction unit determines the sea surface temperature background constraint of the current year, projects the grid-point precipitation prediction results of a certain climate sub-region, a certain reporting date, and a certain forecast date onto the second CDF curve of the same climate sub-region, the same sea surface temperature background constraint, the same reporting date, and the same forecast date as in step S4, and obtains the probability density percentile position corresponding to the precipitation prediction result; maps the probability density percentile position to the first CDF curve of the same climate sub-region, the same sea surface temperature background constraint, and the same date as the forecast date, and obtains the reanalysis precipitation value corresponding to the probability density percentile position. This reanalysis precipitation value is used as the corrected prediction result of the certain reporting date and the forecast date.

[0042] The beneficial effects of this invention are as follows:

[0043] 1. This invention effectively utilizes the scientific support that summer precipitation in my country exhibits significant regional differences due to the southward-to-northward movement of the rain belt, as well as the main characteristic that summer precipitation in my country is significantly affected by the spring Indian Ocean Basin Uniform Mode (IOBM). It designs technical methods and achieves good correction results. This demonstrates the importance of considering regional differences in precipitation and sea surface temperature background constraints in the model error correction of daily summer precipitation in my country predicted by sub-seasonal models, and the effectiveness of scientific methods in guiding technical improvements.

[0044] 2. It solves the problem that the CDF-QM correction method is not applicable due to the significant regional differences in daily precipitation in summer in my country, the tendency for overfitting when modeling single points, and the lack of physical constraints.

[0045] 3. A CDF model is jointly established using gridded data from regions with relatively consistent temporal variations in precipitation to avoid overfitting. Furthermore, different sea surface temperature backgrounds are considered for physical constraints. This model is then used to correct the prediction results for all gridded points within this region. Additionally, an unequal interval setting (dense first, then sparse) is adopted in the CDF model establishment to effectively utilize physically meaningful predictability sources to modify the cumulative probability density correction method of percentile mapping. This corrects the errors in the sub-seasonal model prediction results of daily precipitation in summer in my country, thereby improving my country's summer sub-seasonal precipitation prediction capabilities.

[0046] 4. The method and system of this invention can be extended to other regions outside of my country, and have wider applicability. Attached Figure Description

[0047] Figure 1 The diagram shown is a flowchart of a subseasonal model precipitation correction method based on climate zoning and sea surface temperature background constraints according to an embodiment of the present invention.

[0048] Figure 2 The figure shows the verification score of the error correction effect of the 15-30 day daily precipitation directly output by the CMA-CPSv3 subseasonal model in the embodiment; (a) PC, (b) ETS, and (c) HSS are the forecast accuracy and fairness skill scores, with larger values ​​indicating higher forecast skill; (d) BIAS is the error score, with smaller values ​​indicating higher forecast skill; where Origin is the original result directly output by the subseasonal model, CM1 is the correction result of the cumulative probability density method without considering climate zoning and sea surface temperature background constraints, and CM2 is the correction result of the cumulative probability density method considering climate zoning and sea surface temperature background constraints.

[0049] Figure 3The diagram shown illustrates the real-time prediction model matching and prediction result correction process in steps S5 and S6 of the embodiment. Detailed Implementation

[0050] The specific embodiments of the present invention will be described in detail below with reference to the accompanying drawings. It should be noted that the technical features or combinations of technical features described in the following embodiments should not be considered in isolation, but can be combined with each other to achieve better technical effects.

[0051] like Figure 1 As shown, an embodiment of the present invention provides a method for correcting precipitation in a subseasonal model based on climate zoning and sea surface temperature background constraints, comprising:

[0052] S1. Climate Zoning: The area to be corrected is gridded at a certain resolution; based on the daily characteristics of summer precipitation, the area to be corrected is clustered into several climate sub-regions using reanalysis precipitation data.

[0053] S2. Add sea surface temperature background constraints: Using the sea surface temperature background model, add sea surface temperature background constraints to all the climate sub-regions mentioned in step S1.

[0054] S3. Reanalysis Data Modeling: For each climate sub-region with sea surface temperature background constraints, a sliding window CDF model is performed based on the reanalysis precipitation data of a certain date within a historical period to obtain the first CDF curve of the climate sub-region under the constraints of the sea surface temperature background on that date.

[0055] S4. Modeling historical return data: For each climate sub-region with sea surface temperature background constraints, based on the sub-seasonal model, a sliding window CDF model is performed on all historical return data of the model on a forecast date of a certain reporting date to obtain the second CDF curve of the climate sub-region under the constraints of the reporting date, the forecast date, and the sea surface temperature background.

[0056] S5. Real-time prediction model matching: Determine the sea surface temperature background constraint of the current year, and based on the sub-seasonal model, project the grid-point precipitation prediction results of a certain climate sub-region, a certain reporting date, and a certain forecast date onto the second CDF curve of the same climate sub-region, the same sea surface temperature background constraint, the same reporting date, and the same forecast date as in step S4, to obtain the probability density percentile position corresponding to the precipitation prediction results.

[0057] S6. Prediction result correction: The probability density percentile position obtained in step S5 is mapped to the first CDF curve with the same climate sub-region, the same sea surface temperature background constraint, and the same date as the forecast date obtained in step S3, to obtain the reanalysis precipitation value corresponding to the probability density percentile position. This reanalysis precipitation value is used as the corrected prediction result for the forecast date of the certain reporting date.

[0058] In one specific embodiment, the method further includes: S7, repeating steps S5 and S6, performing real-time prediction model matching and prediction result correction for each forecast day of a certain reporting date, to obtain the corrected prediction results for all forecast days of a certain reporting date.

[0059] In one specific embodiment, the region to be corrected is China. When zoning the climate in S1, based on the daily variation characteristics of precipitation in summer (May-September) in my country, the K-means clustering method is used to divide the 25km resolution grid precipitation reanalysis data across the country into regions (10-20 types of regions are acceptable across the country).

[0060] In one specific embodiment, in S2, the sea surface temperature background model adopts the Indian Ocean Basin Consistent Model (IOBM), which distinguishes between reanalysis precipitation and subseasonal model historical return precipitation in each climatic subregion for years with positive anomalies (years with IOBM indices greater than 1.5 times the standard deviation), negative anomalies (years with IOBM indices less than -1.5 times the standard deviation), and normal years (years with IOBM indices within ±1.5 times the standard deviation).

[0061] In a specific embodiment, in S3, for all grid points in each climate sub-region on a certain day (e.g., May 1, May 2, ..., September 29, September 30), CDF modeling is performed on the precipitation reanalysis data of each climate sub-region within a historical period (currently from 1991 to 2020) according to the positive, negative, and normal years of the IOBM index. A 15-day sliding window (with May 1 as the target day, the sliding window includes April 24 to April 30 and May 1 to May 8) is then performed to obtain the first CDF curve of each climate sub-region under the constraints of the sea surface temperature background on that certain day (any combination of the three factors of climate sub-region, date, and sea surface temperature background constraint corresponds to a first CDF curve).

[0062] In one specific embodiment, in S4, based on the sub-seasonal model, starting from a certain day (e.g., May 1st, May 2nd, ..., September 29th, September 30th), all historical returns of the model are reported for the next x-day (e.g., the prediction period of the CMA-CPSv3 sub-seasonal model is 0-60 days, so x ranges from 0-60). (The historical return periods for different models are different, so...)

[0063] Taking the CMA-CPSv3 subseasonal model as an example, its 2024 forecast results correspond to the historical return period from 2009 to 2023. Based on the IOBM index for positive, negative, and normal years, CDF modeling is performed on all grid points within each climate subregion using a 15-day sliding window (the target day of the sliding window refers to the forecast date, i.e., the aforementioned future x-th day; for example, if the forecast start date is May 1st and the target date is May 2nd, then the sliding window includes precipitation values ​​for April 25th (forecasts started on April 24th), April 26th (forecasts started on April 25th), ..., May 8th (forecasts started on May 7th), and May 9th (forecasts started on May 8th). This yields the second CDF curve for each grid point in the climate subregion under the constraints of the forecast start date, the forecast date, and the sea surface temperature background. For each forecast start date, any combination of the three factors—climate subregion, forecast date, and sea surface temperature background constraint—corresponds to a second CDF curve.

[0064] In a specific embodiment, in S5, the sea surface temperature background of the current year is determined, and the precipitation prediction results of the sub-seasonal model at each grid point from a certain reporting date (e.g., May 1, May 2, ..., September 29, September 30) to the next x-day (e.g., the prediction lead time of the CMA-CPSv3 sub-seasonal model is 0-60 days, so the value of x is 0-60) are projected onto the second CDF curve of the CDF model established in step S4 under the same sea surface temperature background constraint, the same reporting date, the same forecast date, and the same climate sub-region, to determine the percentile position of the predicted precipitation in the second CDF curve.

[0065] In a specific embodiment, in S6, in step S3, the first CDF curve of the CDF model within the same climatic sub-region under the sea surface temperature background constraint of the grid point where the forecast day was corrected in step S5 is located is found, the reanalysis precipitation value corresponding to the percentile position obtained in step S5 in the first CDF curve is determined, and the reanalysis precipitation value is used as the correction result to replace the sub-seasonal model prediction result corrected in step S5.

[0066] Figure 3 In a specific embodiment, the real-time prediction model matching and prediction result correction process in steps S5 and S6 is given. The left image is the second CDF curve, the right image is the first CDF curve, and the predicted precipitation value X is given. f (d) Find the corresponding percentage P on the second CDF curve. c (X f (d)), according to this P c (X f (d) Find the corresponding reanalysis precipitation value X on the first CDF curve. fcorr (d) The reanalysis of precipitation value X fcorr(d) is the final corrected prediction result.

[0067] Repeat steps S3 and S4 to achieve CDF modeling of reanalysis (observation) precipitation for all dates and all sub-regions during the period from May 1 to September 30 under different IOBM sea surface temperature background constraints, as well as CDF modeling of predicted precipitation for all reporting dates, all forecast dates, and all climate sub-regions during the period from May 1 to September 30.

[0068] Repeat steps S5 to S6 to correct the prediction results for all forecast days and spatial grid points of a single reporting date.

[0069] In one specific implementation, in step S1, K-means clustering is used to classify the region to be corrected. All grid points are divided into K groups, and K grid points are randomly selected as the initial cluster centers. Then, the distance (mean squared error) between each grid point and the precipitation time series of other grid points is calculated, and each grid point is assigned to the nearest cluster center. The cluster centers and the grid points assigned to them represent a cluster. Each time a grid point is assigned, the cluster center is recalculated based on all existing grid points in the cluster. This process is repeated until no grid points are reassigned to different clusters, no cluster centers change, and the sum of squared errors is minimized within the same cluster, thus completing the classification of the region to be corrected. Each cluster forms a climate sub-region.

[0070] In one embodiment, the clustering results were divided into 16 categories, that is, into 16 climate sub-regions.

[0071] In one specific implementation, the IOBM index is defined as the anomaly value of the mean sea surface temperature in the 20°S-20°N and 40°E-110°E region after removing the linear trend.

[0072] The standardized calculation method for the IOBM index is as follows:

[0073] here

[0074] Where, x i Represents the IOBM exponent, y i This represents the standardized IOBM index. Let n be the mean of the IOBM index, and n be the number of IOBM indices.

[0075] In one specific embodiment, in step S3, the method for obtaining the first CDF curve is as follows:

[0076] S31. Set the spacing between precipitation groups, wherein the spacing between precipitation groups is gradually set, and the spacing between small precipitation groups is smaller than the spacing between large precipitation groups.

[0077] S32. For the daily precipitation reanalysis values ​​of all grid points in each climate sub-region of the country from May 1 to September 30 according to the clustering results of step S1 within the sliding window of positive, negative and normal years of IOBM index, calculate the cumulative probability density value of precipitation in each precipitation interval according to the set gradual group interval, and obtain the cumulative probability density curve of the entire 0-100 mm range, i.e. the first CDF curve.

[0078] In one specific embodiment, in step S4, the method for obtaining the second CDF curve is as follows:

[0079] S41. Set the interval between precipitation groups, which is the same as in step S31;

[0080] S42. For the daily forecasts from May 1 to September 30 for the next 0-60 days (taking CMA-CPSv3 as an example, the forecast lead time varies for different models), based on the clustering results in step S1, for each grid point in each climate sub-region, calculate the cumulative probability density value of precipitation in each precipitation interval within the entire model historical return period (taking CMA-CPSv3 as an example, its historical return period is the past 15 years with the start year as a reference) according to positive anomaly years, negative anomaly years, and normal years of the IOBM index. This yields the cumulative probability density curve within the range of 0-100 mm, i.e., the second CDF curve.

[0081] In one specific embodiment, the spacing between the precipitation groups is set as follows: 0.1 mm in the range of 0-1 mm, 0.2 mm in the range of 1-10 mm, 0.5 mm in the range of 10-25 mm, 1.0 mm in the range of 25-50 mm, and 2.0 mm in the range of 50-100 mm.

[0082] An embodiment of the present invention provides a subseasonal model precipitation correction system based on climate zoning and sea surface temperature background constraints, comprising:

[0083] The data preprocessing unit is used for standardizing precipitation data and gridding the area to be corrected. The standardization of precipitation data specifically includes: processing the cumulative precipitation reported by the sub-seasonal model and the forecast cumulative precipitation into daily average precipitation; interpolating the reanalysis precipitation data, the daily average precipitation reported by the sub-seasonal model, and the forecast daily average precipitation to grids with the same spatial resolution.

[0084] The climate zoning unit, based on the daily characteristics of summer precipitation, uses the reanalysis precipitation data processed by the data preprocessing unit to cluster and divide the area to be corrected into several climate sub-regions.

[0085] The sea surface temperature background constraint unit uses the sea surface temperature background model to add sea surface temperature background constraints to all the climate sub-regions.

[0086] The data modeling unit then performs sliding window CDF modeling on the reanalysis precipitation data of each sub-region with sea surface temperature background constraints on a certain date within a historical period, to obtain the first CDF curve of the climate sub-region under the constraints of sea surface temperature background on that date.

[0087] The modeling unit for historical return data of the model performs sliding window CDF modeling on all historical return data of the model on a forecast date for a certain forecast date based on the sub-seasonal model for each climate sub-region with sea surface temperature background constraints, and obtains the second CDF curve of the climate sub-region under the constraints of the forecast date, the forecast date, and the sea surface temperature background.

[0088] The real-time model matching and prediction result correction unit determines the sea surface temperature background constraint of the current year, projects the grid-point precipitation prediction results of a certain climate sub-region, a certain reporting date, and a certain forecast date onto the second CDF curve of the same climate sub-region, the same sea surface temperature background constraint, the same reporting date, and the same forecast date as in step S4, and obtains the probability density percentile position corresponding to the precipitation prediction result; maps the probability density percentile position to the first CDF curve of the same climate sub-region, the same sea surface temperature background constraint, and the same date as the forecast date, and obtains the reanalysis precipitation value corresponding to the probability density percentile position. This reanalysis precipitation value is used as the corrected prediction result of the certain reporting date and the forecast date.

[0089] Example 1

[0090] The error correction effect of the CMA-CPSv3 subseasonal model on the prediction results of the summer precipitation peak period from 2021 to 2023, 15 to 30 days in advance.

[0091] ① Data standardization and grid partitioning

[0092] The cumulative precipitation from the CMA-CPSv3 model's historical returns for the summers of 2008 to 2020, and its real-time forecasts for the summers of 2021 to 2023, were processed into daily average precipitation. The reanalysis precipitation, along with the model's historical returns and forecasts, were interpolated to a 25km resolution grid. Based on the interannual variation characteristics of the summer reanalysis precipitation from 1991 to 2020, the K-means method was used to divide the precipitation grid across the country into 16 climatic sub-regions.

[0093] Meanwhile, the IOBM index was calculated for spring (March and April average) from 1991 to 2023. Based on this index, the climatic period of the reanalysis data (1991 to 2020) and the historical return period of CMA-CPSv3 (2008-2020) were divided into IOBM index positive anomaly years, IOBM index negative anomaly years, and IOBM index normal years according to a threshold of 1.5 times the standard deviation. The years to be corrected (2021 to 2023) were determined to be IOBM normal background years.

[0094] ② Further analysis of precipitation CDF modeling

[0095] For the summer reanalysis precipitation of positive, negative, and normal years of IOBM index at 25km grid points in different regions obtained in step ①, a 15-day sliding window CDF model was established for each region.

[0096] ③ Sub-seasonal model return precipitation CDF modeling

[0097] For each region, a 15-day sliding window CDF model was established for the historical daily precipitation reports of the IOBM index in different summer reporting time patterns for positive, negative, and normal years using CMA-CPSv3 at 25km grid points in different regions obtained in step ①.

[0098] ④ Error corrections were made to the daily precipitation forecasts for summer in my country from 2021 to 2023, which were made 15-30 days in advance by CMA-CPSv3.

[0099] The daily forecasts of the CMA-CPSv3 subseasonal model, 15 to 30 days in advance, are projected grid by grid onto the corresponding CDF model built in step ③ to obtain the CDF percentile corresponding to the corrected value; then, this percentile is projected onto the corresponding CDF model built in step ② to obtain the corrected value of the forecast result.

[0100] Skill scoring of prediction results before and after correction, such as Figure 2 As shown, (a) PC, (b) ETS, and (c) HSS are the forecast accuracy and fairness skill scores, with higher values ​​indicating higher forecast skill; (d) BIAS is the error score, with lower values ​​indicating higher forecast skill. The test results of multiple indicators all show that the method of the present invention can effectively correct the forecast skill of sub-seasonal models for summer precipitation in my country.

[0101] Example 2

[0102] The CMA-CPSv3 subseasonal model's precipitation forecasts for June 9, 2024, were revised 15 days and 30 days in advance (2024 is a positive anomaly background year for IOBM). The results demonstrate that the method of this invention has a significant correction effect on the location and magnitude of areas with high precipitation values ​​predicted by CMA-CPSv3 in my country.

[0103] This invention effectively utilizes the scientific support that summer precipitation in my country exhibits significant regional differences due to the southward-to-northward movement of the rain belt, as well as the main characteristic that summer precipitation in my country is significantly affected by the spring Indian Ocean Basin Uniform Mode (IOBM), and designs technical methods that have achieved good correction results.

[0104] While several embodiments of the present invention have been provided herein, those skilled in the art should understand that modifications can be made to these embodiments without departing from the spirit of the invention. The above embodiments are merely exemplary and should not be construed as limiting the scope of the invention.

Claims

1. A method for correcting precipitation in a subseasonal model based on climate zoning and sea surface temperature background constraints, characterized in that, The method includes: S1. Climate Zoning: The area to be corrected is gridded at a certain resolution; based on the daily characteristics of summer precipitation, the area to be corrected is clustered into several climate sub-regions using reanalysis precipitation data. S2. Add sea surface temperature background constraints: Using the sea surface temperature background model, add sea surface temperature background constraints to all the climate sub-regions mentioned in step S1. S3. Reanalysis Data Modeling: For each climate sub-region with sea surface temperature background constraints, a sliding window CDF model is performed based on the reanalysis precipitation data of a certain date within a historical period to obtain the first CDF curve of the climate sub-region under the constraints of the sea surface temperature background on that date. S4. Modeling historical return data: For each climate sub-region with sea surface temperature background constraints, based on the sub-seasonal model, a sliding window CDF model is performed on all historical return data of the model on a forecast date of a certain reporting date to obtain the second CDF curve of the climate sub-region under the constraints of the reporting date, the forecast date, and the sea surface temperature background. S5. Real-time prediction model matching: Determine the sea surface temperature background constraint of the current year, and based on the sub-seasonal model, project the grid-point precipitation prediction results of a certain climate sub-region, a certain reporting date, and a certain forecast date onto the second CDF curve of the same climate sub-region, the same sea surface temperature background constraint, the same reporting date, and the same forecast date as in step S4, to obtain the probability density percentile position corresponding to the precipitation prediction results. S6. Prediction result correction: The probability density percentile position obtained in step S5 is mapped to the first CDF curve with the same climate sub-region, the same sea surface temperature background constraint, and the same date as the forecast date obtained in step S3, to obtain the reanalysis precipitation value corresponding to the probability density percentile position. This reanalysis precipitation value is used as the corrected prediction result for the forecast date of the certain reporting date.

2. The subseasonal model precipitation correction method based on climate zoning and sea surface temperature background constraints as described in claim 1, characterized in that, The method further includes: S7. Repeat steps S5 and S6 to perform real-time prediction model matching and prediction result correction for each forecast day of a given forecast date, and obtain the corrected prediction results for all forecast days of a given forecast date.

3. The subseasonal model precipitation correction method based on climate zoning and sea surface temperature background constraints as described in claim 1, characterized in that, In step S1, K-means clustering is used to classify the region to be corrected. All grid points are divided into K groups, and K grid points are randomly selected as the initial cluster centers. Then, the distance between each grid point and the precipitation time series of other grid points is calculated, and each grid point is assigned to the nearest cluster center. The cluster centers and the grid points assigned to them represent a cluster. Each time a grid point is assigned, the cluster center is recalculated based on all existing grid points in the cluster. This process is repeated until no grid points are reassigned to different clusters, no cluster centers change, and the sum of squared errors is minimized within the same cluster, thus completing the classification of the region to be corrected. Each cluster forms a climate sub-region.

4. The subseasonal model precipitation correction method based on climate zoning and sea surface temperature background constraints as described in claim 1, characterized in that, In step S2, the area to be corrected is the entire China region, and the sea surface temperature background model adopts the Indian Ocean Basin Consistent Model (IOBM). Based on the IOBM index, years are divided into normal years, positive anomaly years, and negative anomaly years. Years with an IOBM index greater than 1.5 times the standard deviation are positive anomaly years, years with an IOBM index less than -1.5 times the standard deviation are negative anomaly years, and years with an IOBM index within ±1.5 times the standard deviation are normal years.

5. The subseasonal model precipitation correction method based on climate zoning and sea surface temperature background constraints as described in claim 4, characterized in that, The IOMB index is defined as the anomaly value of the average sea surface temperature in the 20°S-20°N and 40°E-110°E region after removing the linear trend. The standardized calculation method for the IOBM index is as follows: here Where, x i Represents the IOBM exponent, y i This represents the standardized IOBM index. Let n be the mean of the IOBM index, and n be the number of IOBM indices.

6. The subseasonal model precipitation correction method based on climate zoning and sea surface temperature background constraints as described in claim 4, characterized in that, In step S3, the method for obtaining the first CDF curve is as follows: S31. Set the spacing between precipitation groups, wherein the spacing between precipitation groups is gradually set, and the spacing between small precipitation groups is smaller than the spacing between large precipitation groups. S32. For the daily precipitation reanalysis values ​​of all grid points in each climate sub-region of the country from May 1 to September 30 according to the clustering results of step S1 within the sliding window of positive, negative and normal years of IOBM index, calculate the cumulative probability density value of precipitation in each precipitation interval according to the set gradual group interval, and obtain the cumulative probability density curve of the entire 0-100 mm range, i.e. the first CDF curve.

7. The subseasonal model precipitation correction method based on climate zoning and sea surface temperature background constraints as described in claim 6, characterized in that, In step S4, the method for obtaining the second CDF curve is as follows: S41. Set the interval between precipitation groups, which is the same as in step S31; S42. For daily reports from May 1 to September 30, forecast the next 0-60 days nationwide. Based on the clustering results in step S1, for each climatic sub-region, calculate the cumulative probability density value of the model's historical precipitation within each precipitation interval according to the IOBM index for positive, negative, and normal years. This yields the cumulative probability density curve within the range of 0-100 mm, i.e., the second CDF curve.

8. The subseasonal model precipitation correction method based on climate zoning and sea surface temperature background constraints as described in claim 7, characterized in that, The spacing between the precipitation groups is set as follows: 0.1 mm for 0-1 mm, 0.2 mm for 1-10 mm, 0.5 mm for 10-25 mm, 1.0 mm for 25-50 mm, and 2.0 mm for 50-100 mm.

9. The subseasonal model precipitation correction method based on climate zoning and sea surface temperature background constraints as described in claim 1, characterized in that, In step S5, the judgment and calculation are performed one by one for all forecast days and all grid points for each reporting start date.

10. A subseasonal model precipitation correction system based on climate zoning and sea surface temperature background constraints, characterized in that, The system includes: The data preprocessing unit is used for standardizing precipitation data and gridding the area to be corrected. The standardization of precipitation data specifically includes: processing the cumulative precipitation reported by the sub-seasonal model and the forecast cumulative precipitation into daily average precipitation; interpolating the reanalysis precipitation data, the daily average precipitation reported by the sub-seasonal model, and the forecast daily average precipitation to grids with the same spatial resolution. The climate zoning unit, based on the daily characteristics of summer precipitation, uses the reanalysis precipitation data processed by the data preprocessing unit to cluster and divide the area to be corrected into several climate sub-regions. The sea surface temperature background constraint unit uses the sea surface temperature background model to add sea surface temperature background constraints to all the climate sub-regions. The data modeling unit then performs sliding window CDF modeling on the reanalysis precipitation data of each sub-region with sea surface temperature background constraints on a certain date within a historical period, to obtain the first CDF curve of the climate sub-region under the constraints of sea surface temperature background on that date. The modeling unit for historical return data of the model performs sliding window CDF modeling on all historical return data of the model on a forecast date for a certain forecast date based on the sub-seasonal model for each climate sub-region with sea surface temperature background constraints, and obtains the second CDF curve of the climate sub-region under the constraints of the forecast date, the forecast date, and the sea surface temperature background. The real-time model matching and prediction result correction unit determines the sea surface temperature background constraint of the current year, projects the grid-point precipitation prediction results of a certain climate sub-region, a certain reporting date, and a certain forecast date onto the second CDF curve of the same climate sub-region, the same sea surface temperature background constraint, the same reporting date, and the same forecast date as in step S4, and obtains the probability density percentile position corresponding to the precipitation prediction result; maps the probability density percentile position to the first CDF curve of the same climate sub-region, the same sea surface temperature background constraint, and the same date as the forecast date, and obtains the reanalysis precipitation value corresponding to the probability density percentile position. This reanalysis precipitation value is used as the corrected prediction result of the certain reporting date and the forecast date.

Images