CN114677059B - Method and system for comprehensively evaluating precision of inversion precipitation product by integrating time-space indexes - Google Patents
Method and system for comprehensively evaluating precision of inversion precipitation product by integrating time-space indexes Download PDFInfo
- Publication number
- CN114677059B CN114677059B CN202210578132.7A CN202210578132A CN114677059B CN 114677059 B CN114677059 B CN 114677059B CN 202210578132 A CN202210578132 A CN 202210578132A CN 114677059 B CN114677059 B CN 114677059B
- Authority
- CN
- China
- Prior art keywords
- precipitation
- space
- time
- precision
- data
- 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.)
- Active
Links
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q10/00—Administration; Management
- G06Q10/06—Resources, workflows, human or project management; Enterprise or organisation planning; Enterprise or organisation modelling
- G06Q10/063—Operations research, analysis or management
- G06Q10/0639—Performance analysis of employees; Performance analysis of enterprise or organisation operations
- G06Q10/06393—Score-carding, benchmarking or key performance indicator [KPI] analysis
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/22—Matching criteria, e.g. proximity measures
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/24—Classification techniques
- G06F18/241—Classification techniques relating to the classification model, e.g. parametric or non-parametric approaches
-
- Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
- Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
- Y02A—TECHNOLOGIES FOR ADAPTATION TO CLIMATE CHANGE
- Y02A90/00—Technologies having an indirect contribution to adaptation to climate change
- Y02A90/10—Information and communication technologies [ICT] supporting adaptation to climate change, e.g. for weather forecasting or climate simulation
Landscapes
- Engineering & Computer Science (AREA)
- Business, Economics & Management (AREA)
- Human Resources & Organizations (AREA)
- Theoretical Computer Science (AREA)
- Data Mining & Analysis (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Educational Administration (AREA)
- Evolutionary Biology (AREA)
- General Engineering & Computer Science (AREA)
- Bioinformatics & Computational Biology (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Artificial Intelligence (AREA)
- Life Sciences & Earth Sciences (AREA)
- Development Economics (AREA)
- Evolutionary Computation (AREA)
- Economics (AREA)
- Entrepreneurship & Innovation (AREA)
- Strategic Management (AREA)
- Game Theory and Decision Science (AREA)
- Marketing (AREA)
- Operations Research (AREA)
- Quality & Reliability (AREA)
- Tourism & Hospitality (AREA)
- General Business, Economics & Management (AREA)
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
The invention provides a comprehensive evaluation method and a comprehensive evaluation system for the precision of an inversion precipitation product integrating time-space indexes. The method comprises the following steps: and the data acquisition unit collects the high-density ground rainfall station network observation rainfall data in the preset time and space range and guides the ground reference rainfall data into the database. And leading the space-time resolution attribute of the precipitation product to be evaluated into a space-time data preprocessing unit, carrying out space-time aggregation on the precipitation product to be evaluated to enable the space-time resolution to be the same as ground reference precipitation data, leading the precipitation product to be evaluated after preprocessing into a space-time precision evaluating unit, and outputting the precipitation product to be evaluated into a plurality of space-time precision index subsets. And setting weights for each time-space precision index subset, carrying out weighted summation, and finally outputting the sum as a comprehensive index by an output module. According to the method, a series of spatial precision indexes such as precipitation spatial distribution classification identification and spatial structure depiction are added, and the comprehensiveness and effectiveness of precision comment of precipitation inversion products are improved.
Description
Technical Field
The invention relates to G06F: the field of electric digital data processing, in particular to a precipitation product precision evaluation method, and specifically relates to an inversion precipitation product precision comprehensive evaluation method and system integrating time-space indexes.
Background
The rainfall spatial-temporal distribution and the evolution law thereof are deeply known, and the method has important theoretical and practical significance for promoting climate change research, ecological environment protection, water and soil resource optimal allocation, prevention and control of flood and drought disasters and associated geological disasters and the like. In the past decades, with the continuous improvement of sensor upgrading and inversion algorithms, a series of grid precipitation estimation products (GPEs) which are analyzed and output by satellite, radar inversion and weather numerical model re-analysis are developed. These products have shown good potential in meteorological hydrological studies in data-deficient or data-free areas, becoming important alternative data. However, a large number of evaluation studies show that GPEs have certain advantages in representing precipitation spatial distribution, but due to the influence of sensor measurement and inversion algorithms, insufficient spatial and temporal resolution and prominent quantitative errors are main problems affecting effective application of the data, and the errors are different according to different climatic conditions, terrains, geographical positions, spatial and temporal dimensions and the like.
At present, the knowledge about the estimation accuracy of the lattice precipitation estimation product is basically obtained by considering the average time sequence accuracy of precipitation and the spatial distribution thereof by taking the time sequence precipitation at the position of a rain gauge or a grid as a research object. The spatial accuracy indexes of precipitation estimation products such as classification identification, spatial structure depiction and quantitative estimation effect on precipitation spatial distribution are not paid enough attention. However, the spatial accuracy index is also one of the important indexes in the accuracy index of precipitation products, and even on the premise of the same surface rainfall, different precipitation spatial distributions have considerable influence on river and lake water level calculation, flood control and waterlogging removal standard formulation and engineering layout planning in a drainage basin area. Therefore, a system is lacking to solve the above problems.
Disclosure of Invention
The purpose of the invention is as follows: the method for comprehensively evaluating the precision of the inversion precipitation product integrated with the space-time index is provided, and a system for driving the method is further provided so as to solve the problems in the prior art.
In a first aspect, a comprehensive evaluation method for the precision of an inversion precipitation product integrating space-time indexes is provided, and the method comprises the following steps:
step 2, importing the space-time resolution attribute of the precipitation product to be evaluated into a space-time data preprocessing unit, performing space-time polymerization on the precipitation product to be evaluated by the space-time data preprocessing unit to enable the space-time resolution to be the same as ground reference precipitation data, importing the preprocessed precipitation product to be evaluated into a space-time precision evaluating unit, and finally outputting the preprocessed precipitation product to be evaluated into a plurality of space-time precision index subsets;
and 3, setting weights for each space-time precision index subset output by the space-time precision evaluation unit, carrying out weighted summation, and finally outputting the sum as a comprehensive index by an output module, wherein the comprehensive index reflects the space-time estimation precision of the precipitation product.
In a further embodiment of the first aspect, step 1 further comprises:
step 1-1, estimating precipitation of each lattice point by a data acquisition unit on the basis of estimating dry-wet distribution by adopting the two-stage geographical weighted regression method to obtain a ground reference precipitation data set;
step 1-2, randomly classifying the ground rainfall stations by adopting a K-means clustering method, and evaluating the accuracy of ground reference rainfall data space estimation by using a cross validation method.
In a further embodiment of the first aspect, the spatiotemporal data preprocessing unit, when performing the spatiotemporal aggregation on the precipitation product to be evaluated, further comprises the steps of:
step 2-1, integrating the space-time resolution attributes of ground reference precipitation and precipitation products to be evaluated by the space-time data preprocessing unit, and determining a precision evaluation space scale and a precision evaluation time scale;
step 2-2, performing space aggregation and time accumulation on precipitation to be evaluated to obtain a plurality of space-time precision index subsets with the same resolution as ground reference precipitation data;
the space-time precision evaluation unit comprises a time sequence classification identification capability evaluation module, a time sequence quantitative error evaluation module, a space structure similarity evaluation module and different-intensity precipitation precision evaluation modules;
evaluating the time sequence classification identification capability of the precipitation product by the time sequence classification identification capability evaluation module to generate a first preprocessing data subset;
evaluating the time sequence quantitative error of the precipitation product by the time sequence quantitative error evaluation module to generate a second preprocessing data subset;
the spatial structure similarity evaluation module takes precipitation products and ground reference precipitation data as images, introduces a structural similarity index in the image quality field to evaluate the similarity of precipitation spatial structures, and generates a third preprocessing data subset;
the precipitation intensity is divided into a plurality of intervals by the precipitation accuracy evaluation module with different intensities, and the correct identification proportion of precipitation in the intervals is evaluated by adopting the correct identification rate; and evaluating the average error of the precipitation in a plurality of intervals by using the average relative error to generate a fourth preprocessing data subset.
In a further embodiment of the first aspect, the time-series classification discriminative power evaluation module, when evaluating the time-series classification discriminative power of the precipitation product, further comprises the steps of:
calculating the volume hit index VHI of the precipitation product to be evaluated:
calculating the volume false alarm rate VFAR:
calculate volume critical success index VCSI:
in the formula (I), the compound is shown in the specification,
representing the precipitation of the inversion of the ith period of a certain grid;
representing the observed precipitation on the ground in the ith period of a certain grid; n represents the number of time segments; t represents the threshold value of the precipitation event, and T =0.1mm/d is taken in time sequence precision analysis;
the volume hit index VHI is positively correlated with the time sequence classification identification capability, the calculated volume false alarm rate VFAR is negatively correlated with the time sequence classification identification capability, and the calculated volume key success index VCSI is positively correlated with the time sequence classification identification capability.
In a further embodiment of the first aspect, the spatial structure similarity evaluation module calculates a structural similarity index
The calculation process of (c) is as follows:
in the formula (I), the compound is shown in the specification,
representing the inversion precipitation root mean square error in the analysis unit of the preset time period;
representing the root mean square error of the ground observation precipitation of the analysis unit in a preset time period;
representing the covariance of the inversion and the root mean square error of the ground observed precipitation in the analysis unit of the preset time period; the constant c is used for increasing the stability of the calculation formula under the condition that the variability function of the mean value or the variance converges to the X axis, and the precipitation Range is inverted in the sliding window to calculate
。
In a further embodiment of the first aspect, the different-intensity precipitation accuracy assessment module, when performing the assessment, further comprises the steps of:
dividing the precipitation intensity into 8 different intervals by taking 0.1mm/d, 1mm/d, 5mm/d, 10mm/d, 25mm/d, 50mm/d and 100 mm/d;
and (3) evaluating the correct identification proportion of the precipitation events in different intensity intervals by adopting the correct identification rate CIR:
in the formula (I), the compound is shown in the specification,
、
respectively the lower bound and the upper bound of the water strength interval; r is any real number; the other symbols have the same meanings as above; n represents the number of time segments;
representing the precipitation of the inversion of the ith period of a certain grid;
representing the observed precipitation on the ground in the ith period of a certain grid;
using average relative error MAE PI Calculating the average error of the rainfall estimation in each interval:
in the formula (I), the compound is shown in the specification,
representing the precipitation of the inversion of the ith period of a certain grid;
representing the observed precipitation on the ground in the ith period of a certain grid;
、
the lower and upper bounds of the water intensity interval, respectively.
In a further embodiment of the first aspect, step 3 further comprises:
3-1, establishing a relative membership fuzzy evaluation matrix for each time-space precision index, and determining weight;
and 3-2, carrying out weighted summation on each index, and constructing a comprehensive index reflecting the space-time estimation precision of the precipitation product.
The system comprises at least three components of a data acquisition unit, a space-time data preprocessing unit and an output module.
The data acquisition unit is used for collecting high-density ground rainfall station network observation rainfall data in a preset time and preset space range, preparing ground reference rainfall data based on the high-density ground rainfall station network observation rainfall data, and importing the ground reference rainfall data into a database;
the space-time data preprocessing unit is used for receiving and identifying the space-time resolution attribute of the precipitation product to be evaluated, performing space-time aggregation on the precipitation product to be evaluated by the space-time data preprocessing unit to enable the space-time resolution to be the same as ground reference precipitation data, guiding the precipitation product to be evaluated after preprocessing into the space-time precision evaluating unit, and finally outputting the precipitation product to be evaluated into a plurality of space-time precision index subsets;
and the output module is used for setting weight for each space-time precision index subset output by the space-time precision evaluation unit, performing weighted summation and finally outputting a comprehensive index reflecting the space-time estimation precision of precipitation products.
In a further embodiment of the second aspect, the time-space accuracy assessment unit comprises a time-sequence classification identification capability assessment module, a time-sequence quantitative error assessment module, a spatial structure similarity assessment module and a different-intensity precipitation accuracy assessment module;
evaluating the time sequence classification identification capability of the precipitation product by the time sequence classification identification capability evaluation module to generate a first preprocessing data subset;
evaluating the time sequence quantitative error of the precipitation product by the time sequence quantitative error evaluation module to generate a second preprocessing data subset;
the spatial structure similarity evaluation module takes precipitation products and ground reference precipitation data as images, introduces a structural similarity index in the image quality field to evaluate the similarity of precipitation spatial structures, and generates a third preprocessing data subset;
the precipitation intensity is divided into a plurality of intervals by the precipitation accuracy evaluation module with different intensities, and the correct identification proportion of precipitation in the intervals is evaluated by adopting the correct identification rate; and evaluating the average error of the precipitation in a plurality of intervals by using the average relative error to generate a fourth preprocessing data subset.
The technical scheme of the invention has the following advantages:
in the stage of preparing reference rainfall data/aggregating the spatial and temporal resolutions of the rainfall product to be evaluated, high-precision rainfall reference data is prepared by adopting two-stage geographical weighted regression (TGWR) based on ground high-density rainfall station network data, the precision is evaluated by adopting a cross validation method, and simultaneously the spatial and temporal aggregation of the rainfall product to be evaluated is carried out to ensure that the spatial and temporal resolutions of the rainfall product to be evaluated are consistent with the spatial and temporal resolutions of the reference rainfall. In the stage of evaluating the classification identification capability and the timing sequence quantitative error, indexes such as a Volume Hit Index (VHI), a Volume False Alarm Rate (VFAR), a Volume Critical Success Index (VCSI), a conventional Mean Absolute Error (MAE), a root mean square error (SRMSE) and the like are introduced to evaluate the classification identification capability and the timing sequence quantitative error of the precipitation product. In the spatial accuracy evaluation stage, a structural similarity index (S) creatively introduced into the image quality field evaluates the similarity of the spatial structure of the precipitation. And in the stage of evaluating the precipitation precision of different intensity intervals, adopting Correct Identification Rate (CIR) and average relative error (MAE) to respectively evaluate the correct identification proportion of the precipitation of different intensity intervals and the average error of the precipitation of each interval. Generally, the invention provides a set of universal precipitation precision evaluation framework comprising reference precipitation preparation, precipitation space-time resolution aggregation to be evaluated, time sequence precision evaluation, space precision evaluation and precipitation precision evaluation of different strength intervals. The method improves the defect that the original rainfall accuracy assessment method emphasizes time sequence accuracy comment and ignores space accuracy comment, further increases a series of space accuracy indexes such as rainfall space distribution classification identification and space structure depiction, perfects a space-time estimation accuracy index system of a rainfall inversion product, and improves the comprehensiveness and effectiveness of the rainfall inversion product accuracy comment.
Drawings
FIG. 1 is a flow chart of the practice of the present invention.
FIG. 2 shows an embodiment of the present invention: position diagram of the river basin.
Fig. 3 is a view at a in fig. 2: the distribution diagram of the rainfall station above the mussel port sluice hydrological station in the river basin is shown.
FIG. 4 is a scatter plot of 5 precipitation inversion products and baseline precipitation on a 0.25 by 0.25 grid cell in accordance with an embodiment of the present invention.
Detailed Description
In the following description, numerous specific details are set forth in order to provide a more thorough understanding of the present invention. It will be apparent, however, to one skilled in the art, that the present invention may be practiced without one or more of these specific details. In other instances, well-known features have not been described in order to avoid obscuring the invention.
The embodiment provides an inversion precipitation product precision comprehensive evaluation method integrating space-time indexes by utilizing quantitative and volume classification detection indexes, and the region above the mussel port gate hydrological station in the river basin is taken as a specific region for explanation, and the method is shown in fig. 2 and 3 and is an embodiment of the invention: the distribution diagram of the precipitation stations above the mussel port sluice hydrological station in the river basin is shown. The area of the south-north climate transition zone in the east of China in the Huaihe river basin is about 27 km 2 . Tuebieshan and Funishan are distributed in the northwest of the drainage basin, Dabie mountain in the southwest and Yimeng mountain in the northeast, and the rest are wide plains. The average annual precipitation of the Huaihe river basin is about 920mm, the total amount of the rainfall decreases gradually from south to north, the annual distribution is extremely uneven, and the annual precipitation accounts for 50-80% of the annual precipitation in the flood season (6-9 months). The mussel port hydrological station is the main control station for the middle trip of the river. The present example uses 5 precipitation products, MSWEP V2.1, MSWEP V2.2, TRMM 3B 42V 7, CMORPH BLD, ERA5, for space-time accuracy assessment. The method mainly comprises the following steps:
s1): preparing standard precipitation data and evaluating the precision: in view of the outstanding spatial discontinuity problem of short-duration rainfall on a daily scale and the like, two stages are providedGWRMethod Two-stageGWR(TGWR) For estimating spatial distribution of precipitation, i.e. usingGWREstimating drynessAnd on the basis of distribution, estimating the rainfall of each grid point to obtain a ground grid point reference rainfall data set. To verify the validity of the algorithm, use is made ofKThe mean value clustering method randomly divides 539 rainfall stations into 4 types, and the estimation accuracy of the ground rainfall space is evaluated through a cross validation method (namely the spatial distribution modes of 4 modeling stations and inspection stations).
S2): the precipitation product to be evaluated polymerizes in space and time: and (3) integrating the space-time resolution attributes of the ground reference precipitation and 5 precipitation products, and determining a grid with the precision evaluation space scale of 0.25 degrees multiplied by 0.25 degrees and the time scale of day. FIG. 4 is a scatter plot of 5 precipitation inversion products and a baseline precipitation on a 0.25 by 0.25 grid cell, as shown in FIG. 4, according to an embodiment of the present invention.
Therefore, space polymerization and time accumulation are carried out on the MSWEP to obtain 0.25 degrees multiplied by 0.25 degrees daily rainfall from 2006 to 2015. The TRMM 3B 42V 7, CMORPH BLD and ERA5 spatial resolution meet the requirements, and daily precipitation data are obtained by summing in the time dimension.
S3): evaluating the time sequence classification identification capability of precipitation products: table 1 counts the average of the time sequence accuracy of 5 inverted precipitation products across all 0.25 ° grids. As can be seen from the table, the performances of MSWEP V2.2 and V2.1 are basically equivalent, VHI is close to 1, VFAR is less than 0.1, VCSI exceeds 0.9, and the classification identification capability is very excellent; meanwhile, CC reaches 0.87, which indicates that the time interval synchronism of MSWEP and observed precipitation is good, but certain quantitative error exists, and SRMSE is 1.79. The 2 kinds of satellite inversion precipitation slightly underestimate the daily precipitation, but MAE and SRMSE of CMROPH BLD reach MSWEP level, each index is obviously superior to TRMM 3B 42V 7, and TRMM 3B 42V 7 classification identification and quantitative estimation accuracy is even inferior to ERA 5. The VHI, VFAR and VCSI of ERA5 are close to MSWEP V2.2, but the error is more prominent. Overall, the 5 inversion products have better rain/no-rain resolving power, but the quantitative error is not negligible. Consistent with the laws presented by the scatter plots, the 2 MSWEP and CMORPH BLD timing accuracies were comparable and optimal, followed by ERA5, TRMM 3B 42V 7 with the worst overall accuracy.
Table 15 mean values of time sequence accuracy of inverted precipitation products on all 0.25 ° grids
| Metrics | MSWEP V2.2 | MSWEP V2.1 | TRMM 3B42 V7 | CMORPH BLD | ERA5 |
| VHI | 0.99 | 0.98 | 0.91 | 1.00 | 0.99 |
| VFAR | 0.07 | 0.06 | 0.12 | 0.07 | 0.09 |
| VCSI | 0.92 | 0.92 | 0.81 | 0.93 | 0.91 |
| CC | 0.87 | 0.87 | 0.72 | 0.87 | 0.75 |
| ME/(mm/d) | 0.0 | 0.1 | -0.3 | -0.2 | -0.1 |
| MAE/(mm/d) | 1.2 | 1.2 | 2.0 | 1.2 | 1.6 |
| SRMSE | 1.79 | 1.80 | 2.77 | 1.80 | 2.45 |
S4): and (3) evaluating the timing sequence quantitative error of precipitation products: considering the spatial position attribute of the grids, for the classification accuracy index, except TRMM 3B 42V 7, VHIs of other GPEs have higher homogeneity in space, VFAR and VCSI both have the characteristic that the hills in the northwest are slightly lower than those in the south, the spatial distribution of the two MSWEPs is basically consistent, and the VCSI in the south of the CMORPH BLD basin is even higher than the MSWEP. However, due to local topography, TRMM 3B 42V 7 and CMORPH BLD have higher VFAR and significantly lower VCSI at a few points in Funiu mountain and Tumbe mountain. Moreover, large-area water bodies such as high pond lakes, watt port lakes, east city lakes, coke sentry lakes and the like are distributed in lakes in the southeast of the research area, and reflection signals of high-frequency Passive Microwaves (PMW) in the water areas are usually identified as micro precipitation signals by the inversion algorithm of Microwave precipitation, so that the VFAR and VCSI of a few isolated grids have obvious singular phenomena. In terms of quantitative estimation effect, the south part of MAE is higher than the north part, the north part of SRMSE is higher than the south part, which is closely related to the decreasing of precipitation from south to north, the errors of TRMM 3B 42V 7 and ERA5 are more prominent than those of other 3 data, and the grid-connected slice characteristic of CMORPH BLD with higher CC value is stronger than MSWEP.
S5): and (3) evaluating the space precision of precipitation products: the space precision average value of 5 GPEs for estimating daily rainfall in 2006-2015 of the Huaihe river basin is shown in Table 2. In the aspect of classification accuracy, the improvement amplitude of MSWEP V2.2 is obviously larger than that of V2.1, but VHI, VFAR and VCSI of CMROPH BLD are all better than that of MSWEP V2.2, and the capacity of distinguishing rain/no rain precipitation is stronger, the space VFAR of TRMM 3B 42V 7 is smaller than ERA5, and the other 2 indexes are not as good as ERA 5. In the aspect of spatial distribution description, the CMORPH BLD has the strongest characterization capability on the daily precipitation spatial structure of the Huaihe river basin, S is close to 0.5, MSWEP V2.2 is weaker than V2.1, and ERA5 is less than TRMM 3B 42V 7. With respect to quantitative errors, satellite and reanalyzed precipitation underestimates daily precipitation in small steps, while MSWEP is weakly overestimated, the absolute errors of MSWEP V2.2 and CMROPH BLD are both 2.5mm/d and slightly lower than MSWEP V2.1, the relative errors are not, MSWEP V2.1 is lowest, V2.2 times is worse, CMROPH BLD is worse, and the total is higher than the other 2 data. It should be noted that the quantitative error of TRMM 3B 42V 7 is quite prominent, and significantly exceeds ERA 5.
Table 25 GPEs are used for estimating space precision average value of daily rainfall in 2006-2015 year of Huaihe river basin
| Metrics | MSWEP V2.2 | MSWEP V2.1 | TRMM 3B42 V7 | CMORPH BLD | ERA5 |
| VHI | 0.93 | 0.87 | 0.77 | 0.96 | 0.94 |
| VFAR | 0.18 | 0.15 | 0.18 | 0.16 | 0.21 |
| VCSI | 0.79 | 0.75 | 0.66 | 0.82 | 0.76 |
| S | 0.41 | 0.44 | 0.37 | 0.49 | 0.29 |
| ME/(mm/d) | 0.2 | 0.3 | -0.3 | -0.3 | -0.1 |
| MAE/(mm/d) | 2.5 | 2.6 | 4.3 | 2.5 | 3.4 |
| SRMSE | 2.62 | 2.50 | 6.54 | 2.84 | 2.90 |
S6): evaluating the precipitation estimation effect with different intensities: in the rain section range, when the precipitation intensity is smaller than 25mm/d, the identification capability is continuously strengthened along with the increase of the precipitation intensity, after the precipitation intensity exceeds 25mm/d, the classification identification capability tends to be reduced, the identification capability of events larger than 100mm/d is greatly reduced, and the CIRs of other 4 GPEs except ERA5 are about 0.2. MSWEP V2.2 has a greater capacity to capture rain events than V2.1, but weaker than CMORH BLD, the difference tending to decrease as the intensity of precipitation increases. When the precipitation intensity is less than 25mm/d, the CIR of TRMM 3B 42V 7 is less than ERA5, and after reaching 100mm/d, the TRMM 3B 42V 7 is higher than all other products. In terms of quantitative error, GPEs exhibit MAE PI More prominent with increasing precipitation intensity, 2 MSWEP's are substantially equivalent to CMORPH BLD error level when precipitation intensity is less than 50mm/d, and after exceeding the intensity, MSWEP's are slightly greater than CMORPH BLD, and MAE of all products PI The amplification is obviously larger, and the estimation error of the strength exceeding 100mm/d is reachedAbove 45mm/d, ERA5 even exceeds 70 mm/d.
By adopting the technical scheme disclosed by the invention, the following beneficial effects are obtained: the traditional precision evaluation method for the precipitation inversion product mostly takes the time sequence precipitation at the position of a rain gauge or a grid as a research object and obtains the precision by evaluating the average time sequence precision and the spatial distribution of the precipitation, but the traditional method generally rarely touches the classification identification, spatial structure depiction and quantitative estimation effects of the precipitation estimation product on the spatial distribution of the precipitation. On the basis of an original evaluation index system, the method firstly further subdivides a time sequence precision index into time sequence classification identification capability and time sequence quantitative errors, and simultaneously further introduces a structural similarity index in the field of image quality as a precision index for describing the space description capability of the time sequence classification identification capability and the time sequence quantitative errors, and also provides a precision evaluation method for different precipitation intensity intervals, so that the description capability of the precipitation product estimation effect is enhanced, and the method system in the field of precipitation product precision evaluation is expanded. In a specific embodiment, it is found that the new evaluation method system can obtain a conclusion that is not completely consistent with the original precision evaluation method, that is: the precipitation product (in this example, MSWEP) considered as the optimal in the original accuracy evaluation system lacks consideration on the autocorrelation of the local precipitation space due to the algorithm using the isolated grid as an analysis unit, is not good in characterization capability of the precipitation space distribution, is not like other precipitation products such as CMOPRH BLD, and shows a certain disadvantage in the aspect of representing the effect of high-intensity precipitation. Obviously, the provided rainfall precision evaluation method can further deepen the objective knowledge of the rainfall product estimation effect, and has important scientific and practical values for properly selecting rainfall data according to different climate areas and the meteorological hydrological research purpose.
The above embodiments mainly relate to geographical information, rainfall station observation and 5 kinds of grid precipitation products in the aspect of information utilization. It should be noted that the method has strong expansibility, the quantity of integratable information is not limited to the method, and other remote sensing inversion and the precision evaluation of the reanalysis aquatic product can adopt the method.
As noted above, while the present invention has been shown and described with reference to certain preferred embodiments, it is not to be construed as limited thereto. Various changes in form and detail may be made therein without departing from the spirit and scope of the invention as defined by the appended claims.
Claims (2)
1. The method for comprehensively evaluating the precision of an inverted precipitation product integrated with a time-space index is characterized by comprising the following steps of:
step 1, collecting high-density ground rainfall station network observation rainfall data in a preset time and preset space range by a data acquisition unit, preparing ground reference rainfall data based on the high-density ground rainfall station network observation rainfall data, and importing the ground reference rainfall data into a database;
step 1-1, estimating precipitation of each lattice point by a data acquisition unit on the basis of estimating dry-wet distribution by adopting a two-stage geographical weighted regression method to obtain a ground reference precipitation data set;
step 1-2, randomly classifying the ground rainfall stations by adopting a K-means clustering method, and evaluating the accuracy of ground reference rainfall data space estimation by using a cross validation method;
step 2, leading the space-time resolution attribute of the precipitation product to be evaluated into a space-time data preprocessing unit, carrying out space-time polymerization on the precipitation product to be evaluated by the space-time data preprocessing unit to enable the space-time resolution to be the same as ground reference precipitation data, leading the precipitation product to be evaluated after preprocessing into a space-time precision evaluating unit, and finally outputting the precipitation product to be evaluated into a plurality of space-time precision index subsets;
step 2-1, integrating the space-time resolution attributes of ground reference precipitation and precipitation products to be evaluated by the space-time data preprocessing unit, and determining a precision evaluation space scale and a precision evaluation time scale;
step 2-2, performing space aggregation and time accumulation on precipitation to be evaluated to obtain a plurality of space-time precision index subsets with the same resolution as ground reference precipitation data;
the space-time precision evaluation unit comprises a time sequence classification identification capability evaluation module, a time sequence quantitative error evaluation module, a space structure similarity evaluation module and different-intensity precipitation precision evaluation modules;
evaluating the time sequence classification identification capability of the precipitation product by the time sequence classification identification capability evaluation module to generate a first preprocessing data subset;
calculating the volume hit index VHI of the precipitation product to be evaluated:
calculating the volume false alarm rate VFAR:
calculate volume key success index, VCSI:
in the formula (I), the compound is shown in the specification,
representing the precipitation of the inversion of the ith period of a certain grid;
representing the observed precipitation on the ground in the ith period of a certain grid; n represents the number of time segments; t represents the threshold value of a precipitation event, and T =0.1mm/d in the time sequence accuracy analysis;
the volume hit index VHI is positively correlated with the time sequence classification identification capability, the volume false alarm rate VFAR is negatively correlated with the time sequence classification identification capability, and the volume key success index VCSI is positively correlated with the time sequence classification identification capability;
evaluating the timing quantitative error of the precipitation product by the timing quantitative error evaluation module to generate a second preprocessing data subset;
the spatial structure similarity evaluation module takes precipitation products and ground reference precipitation data as images, introduces a structural similarity index in the image quality field to evaluate the similarity of precipitation spatial structures, and generates a third preprocessing data subset;
the spatial structure similarity evaluation module calculates a structural similarity index
The calculation process of (2) is as follows:
in the formula (I), the compound is shown in the specification,
representing the inversion of the precipitation root mean square error in the analysis unit in a predetermined time period;
representing the root mean square error of the ground observation precipitation of the analysis unit in a preset time period;
representing the covariance of the inversion and the root mean square error of the ground observed precipitation in the analysis unit of the preset time period; the constant c is used for increasing the stability of the calculation formula under the condition that the variability function of the mean value or the variance converges to the X axis, and the precipitation Range is inverted in the sliding window to calculate
；
The different-intensity precipitation precision evaluation module divides the precipitation intensity into a plurality of intervals, and evaluates the correct identification proportion of precipitation in the intervals by adopting the correct identification rate; evaluating the average error of the precipitation in a plurality of intervals by adopting the average relative error to generate a fourth preprocessing data subset;
the different-intensity precipitation precision evaluation module further comprises the following steps when the evaluation is executed:
dividing the precipitation intensity into 8 different intervals by taking 0.1mm/d, 1mm/d, 5mm/d, 10mm/d, 25mm/d, 50mm/d and 100 mm/d;
and (3) evaluating the correct identification proportion of the precipitation events in different intensity intervals by adopting the correct identification rate CIR:
wherein R is any real number; the other symbols have the same meanings as above; n represents the number of time segments;
using average relative error MAE PI Calculating the average error of the rainfall estimation in each interval:
in the formula (I), the compound is shown in the specification,
、
the lower and upper bounds of the water strength interval are respectively;
and 3, setting weights for each space-time precision index subset output by the space-time precision evaluation unit, carrying out weighted summation, and finally outputting the sum as a comprehensive index by an output module, wherein the comprehensive index reflects the space-time estimation precision of the precipitation product.
2. The evaluation method according to claim 1, wherein step 3 further comprises:
3-1, establishing a relative membership fuzzy evaluation matrix for each time-space precision index, and determining weight;
and 3-2, carrying out weighted summation on each index, and constructing a comprehensive index reflecting the space-time estimation precision of the precipitation product.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202210578132.7A CN114677059B (en) | 2022-05-26 | 2022-05-26 | Method and system for comprehensively evaluating precision of inversion precipitation product by integrating time-space indexes |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202210578132.7A CN114677059B (en) | 2022-05-26 | 2022-05-26 | Method and system for comprehensively evaluating precision of inversion precipitation product by integrating time-space indexes |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN114677059A CN114677059A (en) | 2022-06-28 |
| CN114677059B true CN114677059B (en) | 2022-08-23 |
Family
ID=82080748
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202210578132.7A Active CN114677059B (en) | 2022-05-26 | 2022-05-26 | Method and system for comprehensively evaluating precision of inversion precipitation product by integrating time-space indexes |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN114677059B (en) |
Families Citing this family (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN115983511B (en) * | 2023-03-22 | 2023-05-23 | 水利部交通运输部国家能源局南京水利科学研究院 | Precipitation prediction method and system based on improved statistical downscaling method |
| CN116740473B (en) * | 2023-08-10 | 2024-01-09 | 中国水产科学研究院南海水产研究所 | Automatic sorting method and system for fish catch based on machine vision |
Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN107316095A (en) * | 2016-09-23 | 2017-11-03 | 武汉大学 | A kind of region meteorological drought grade prediction technique for coupling multi-source data |
| CN108961402A (en) * | 2018-06-19 | 2018-12-07 | 河海大学 | Space-time precision calibration method of the multi-satellite remote sensing precipitation inverting in large scale complexity basin |
| CN109946762A (en) * | 2019-03-06 | 2019-06-28 | 重庆邮电大学移通学院 | A kind of method and system based on probability distribution Short-term Forecast precipitation |
Family Cites Families (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN104392097A (en) * | 2014-10-24 | 2015-03-04 | 封国林 | Seasonal precipitation analogue prediction method based on seasonal prediction mode |
| US10107939B2 (en) * | 2016-05-20 | 2018-10-23 | The Climate Corporation | Radar based precipitation estimates using spatiotemporal interpolation |
| CN111104640A (en) * | 2019-11-14 | 2020-05-05 | 河海大学 | Rainfall observation and evaluation method and system based on analytic hierarchy process |
| CN111797131B (en) * | 2020-06-09 | 2024-02-20 | 武汉大学 | Extreme precipitation area frequency analysis method based on remote sensing precipitation product |
| CN113050195B (en) * | 2021-02-07 | 2022-07-15 | 国家气象中心(中央气象台) | Hourly resolution precipitation process identification method |
| CN112800634B (en) * | 2021-04-07 | 2021-06-25 | 水利部交通运输部国家能源局南京水利科学研究院 | Rainfall estimation method and system coupling dry-wet state identification and multi-source information fusion |
-
2022
- 2022-05-26 CN CN202210578132.7A patent/CN114677059B/en active Active
Patent Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN107316095A (en) * | 2016-09-23 | 2017-11-03 | 武汉大学 | A kind of region meteorological drought grade prediction technique for coupling multi-source data |
| CN108961402A (en) * | 2018-06-19 | 2018-12-07 | 河海大学 | Space-time precision calibration method of the multi-satellite remote sensing precipitation inverting in large scale complexity basin |
| CN109946762A (en) * | 2019-03-06 | 2019-06-28 | 重庆邮电大学移通学院 | A kind of method and system based on probability distribution Short-term Forecast precipitation |
Also Published As
| Publication number | Publication date |
|---|---|
| CN114677059A (en) | 2022-06-28 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN114677059B (en) | Method and system for comprehensively evaluating precision of inversion precipitation product by integrating time-space indexes | |
| Carpenter et al. | Intercomparison of lumped versus distributed hydrologic model ensemble simulations on operational forecast scales | |
| CN112800634B (en) | Rainfall estimation method and system coupling dry-wet state identification and multi-source information fusion | |
| Xie et al. | Practical thresholds for separating erosive and non–erosive storms | |
| CN113887972A (en) | Comprehensive drought monitoring and evaluating method based on hydrological process | |
| Zhao et al. | Using satellite remote sensing to understand maize yield gaps in the North China Plain | |
| Houston et al. | Thunderstorm Observation by Radar (ThOR): An algorithm to develop a climatology of thunderstorms | |
| CN105678047A (en) | Wind field characterization method with empirical mode decomposition noise reduction and complex network analysis combined | |
| Satalino et al. | Wheat crop mapping by using ASAR AP data | |
| Raj et al. | Towards evaluating gully erosion volume and erosion rates in the Chambal badlands, Central India | |
| Azizi Mobaser et al. | Evaluating the performance of Era-5 Re-analysis data in estimating daily and monthly precipitation, Case Study; Ardabil Province | |
| Haddjeri et al. | Exploring the sensitivity to precipitation, blowing snow, and horizontal resolution of the spatial distribution of simulated snow cover | |
| Mohtar et al. | Rainfall erosivity estimation for Northern and Southern peninsular Malaysia using Fourneir indexes | |
| Zhang et al. | Parameter calibration and uncertainty estimation of a simple rainfall-runoff model in two case studies | |
| CN109948175B (en) | Satellite remote sensing albedo missing value inversion method based on meteorological data | |
| Wang et al. | Early warning of debris flow using optimized self-organizing feature mapping network | |
| Araújo | Entropy‐based equation to assess hillslope sediment production | |
| Mondal et al. | Land use/Land cover changes in Hugli Estuary using Fuzzy CMean algorithm | |
| Zhou et al. | Daily rainfall model to merge TRMM and ground based observations for rainfall estimations | |
| Thakural et al. | Trend analysis of rainfall for the Chaliyar Basin, South India | |
| Choi et al. | Storm identification and tracking algorithm for modeling of rainfall fields using 1-h NEXRAD rainfall data in Texas | |
| CN111476434A (en) | GIS-based soil heavy metal fractal dimension spatial variation analysis method | |
| Jeewanthi et al. | Appropriate conventional methods for estimating missing precipitation values in Sri Lanka | |
| CN116679357B (en) | Method for carrying out water balance treatment on radar quantitative precipitation estimation | |
| Abdulrahman et al. | REASSESSMENT OF SCS-CN INITIAL ABSTRACTION RATIO BASED ON RAINFALL-RUNOFF EVENT ANALYSIS AND SLOPE-ADJUSTED CN IN A SEMIARID CLIMATE OF HALABJA GOVERNORATE |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |