Search

CN-122020551-A - Triple collocation-based transpiration-transpiration ratio data set fusion method, system, equipment, medium and product

CN122020551ACN 122020551 ACN122020551 ACN 122020551ACN-122020551-A

Abstract

The application discloses a method, a system, equipment, a medium and a product for fusing a transpiration and transpiration ratio data set based on triple collocation, and relates to the field of ecological hydrology. The method comprises the steps of obtaining three sets of transpiration T and transpiration ET data sets of different sources, preprocessing the three sets of transpiration T and transpiration ET data sets to obtain three preprocessed T and ET data sets, respectively calculating error variances of T and ET at each data source by using a triple collocation method for the three preprocessed T and ET data sets, distributing weights of T and ET at each data source according to the error variances of T and ET at each data source, carrying out fusion calculation according to the weights of T and ET at each data source to obtain fused T and ET, and calculating the ratio of the fused T and ET to obtain a fused T/ET data set. The application can effectively solve the problem of systematic deviation of multi-source T/ET data, remarkably improve estimation precision and enhance adaptability to regional heterogeneity.

Inventors

  • HU ZHAOYONG
  • LIN SHAN
  • HUANG KEWEI
  • ZHAO QINGMEI
  • SUN JUYING
  • WANG GENXU
  • SUN SHOUQIN

Assignees

  • 四川大学

Dates

Publication Date
20260512
Application Date
20260202

Claims (10)

  1. 1. The method for fusing the data sets of the transpiration and transpiration ratio based on triple collocation is characterized by comprising the following steps of: Three sets of transpiration T and transpiration ET data sets with different sources are obtained and preprocessed, so that three preprocessed T and ET data sets are obtained; For the three preprocessed T and ET data sets, respectively calculating error variances of T and ET in each data source by using a triple collocation method; distributing weights of T and ET in each data source according to the error variance of T and ET in each data source; Fusion calculation is carried out according to the weights of the T and the ET in each data source, and T and ET after fusion are obtained; And calculating the ratio of the T to the ET after fusion to obtain a T/ET data set after fusion.
  2. 2. The triple collocation-based transpiration/transpiration ratio dataset fusion method according to claim 1, wherein the three sets of transpiration T and transpiration ET datasets from different sources are obtained and preprocessed, and three preprocessed T and ET datasets are obtained, specifically comprising: Three sets of T and ET data sets from different sources are obtained, space-time normalization processing is carried out on the data sets, the data sets are spatially unified into the same resolution, time scales are aligned, missing values are processed, a space mask is generated, only pixels with effective three source data are reserved, three preprocessed T and ET data sets are obtained, and the data of the three preprocessed data sources are respectively recorded as 、 、 And 、 、 。
  3. 3. The triple collocation-based transpiration/transpiration ratio dataset fusion method according to claim 2, wherein for the three preprocessed T and ET datasets, the triple collocation method is used to calculate the error variance of T and ET at each data source, respectively, specifically comprising: For data of three data sources after Tpre-processing 、 、 Using the formula Calculating signal variance of T Wherein Representation of And Is a covariance of (2); Representation of And Is a covariance of (2); Representation of And Is a covariance of (2); Using the formula Calculating T at each data source 、 、 Error variance of (2) 、 、 Wherein 、 、 Respectively represent 、 、 Is a variance of (2); data for three data sources after ET preprocessing 、 、 Using the formula Calculating signal variance of ET Wherein Representation E And Is a covariance of (2); Representation of And Is a covariance of (2); Representation of And Is a covariance of (2); Using the formula Computing ET at each data source 、E 、 Error variance of (2) 、 、 Wherein 、 、 Respectively represent 、 、 Is a variance of (c).
  4. 4. A triple collocation-based transpiration/transpiration ratio dataset fusion method as recited in claim 3, wherein said assigning weights of T and ET at each data source according to the error variances of T and ET at each data source specifically comprises: according to T at each data source 、 、 Error variance of (2) 、 、 Using the formula Distributing T among data sources 、 、 Weights of (2) 、 、 Wherein 、 、 Respectively is 、 、 In the picture element A value at; the number of pixels for T; According to ET at each data source 、 、 Error variance of (2) 、 、 Using the formula Distributing ET among data sources 、E 、E Weights of (2) 、 、 Wherein 、 、 Respectively is 、 、 In the picture element A value at; is the number of picture elements of ET.
  5. 5. The triple collocation-based transpiration/transpiration ratio dataset fusion method of claim 4, wherein the fusion calculation is performed according to weights of the T and the ET in each data source to obtain the fused T and ET, and the method specifically comprises the following steps: according to T at each data source 、 、 Weights of (2) 、 、 Using the formula Performing fusion calculation to obtain fused T, and marking as ; According to ET at each data source 、E 、E Weights of (2) 、 、 Using the formula Performing fusion calculation to obtain fused ET, and recording as 。
  6. 6. The triple collocation-based transpiration/transpiration ratio dataset fusion method of claim 5, wherein the calculating the ratio of the fused T to ET to obtain the fused T/ET dataset specifically comprises: Calculating the ratio of T to ET after fusion And obtaining a fused T/ET data set.
  7. 7. A triple collocation-based transpiration/transpiration ratio dataset fusion system, comprising: the data set acquisition module is used for acquiring three sets of transpiration T and transpiration ET data sets from different sources and preprocessing the three sets of transpiration T and transpiration ET data sets to obtain three preprocessed T and ET data sets; the triple collocation calculation module is used for calculating error variances of the T and the ET in each data source respectively by using a triple collocation method for the three preprocessed T and ET data sets; the weight distribution module is used for distributing weights of T and ET in each data source according to the error variances of the T and ET in each data source; the fusion calculation module is used for carrying out fusion calculation according to the weights of the T and the ET in each data source to obtain the T and the ET after fusion; and the ratio calculating module is used for calculating the ratio of the T to the ET after fusion to obtain a T/ET data set after fusion.
  8. 8. A computer device comprising a memory, a processor and a computer program stored on the memory and executable on the processor, wherein the processor executes the computer program to implement the triple collocation based transpiration to transpiration ratio dataset fusion method of any of claims 1 to 6.
  9. 9. A computer readable storage medium having stored thereon a computer program, wherein the computer program when executed by a processor implements the triple collocation based transpiration to transpiration ratio dataset fusion method of any of claims 1 to 6.
  10. 10. A computer program product comprising a computer program, characterized in that the computer program, when executed by a processor, implements the triple collocation based transpiration to transpiration ratio dataset fusion method as claimed in any of claims 1 to 6.

Description

Triple collocation-based transpiration-transpiration ratio data set fusion method, system, equipment, medium and product Technical Field The application relates to the technical field of ecological hydrology, in particular to a method, a system, equipment, a medium and a product for fusing a data set of transpiration and transpiration ratio based on triple collocation. Background The ratio T/ET of transpiration (Transpiration, T) to transpiration (Evapotranspiration, ET) is an important biophysical parameter measuring the strength of the interaction of the land and gas, and is widely used for evaluating the vegetation water utilization efficiency, the ecosystem carbohydrate coupling process and the water resource allocation pattern. Current research on T/ET is mainly dependent on methods such as flux observation, remote sensing inversion, ecological model simulation, and atmospheric re-analysis. Wherein flux observation can provide more accurate local T/ET measurements, but is limited by the sparse number of sites and the short time series, which makes it difficult to support long-term dynamic assessment of regional or even national dimensions. The remote sensing inversion, ecological model and analysis method has wide space-time coverage capability and has been widely used for large-scale T/ET estimation, however, due to the fact that input variables, physical assumptions and algorithm structures are different, obvious inconsistency and systematic deviation exist among different products, and application reliability of the method in decision support and scientific research is further limited. To make up for the deficiency of a single data source, fusing multi-source T/ET products becomes a key way to improve estimation accuracy. The traditional fusion method directly carries out equal-weight superposition on multi-source data, fails to identify and correct systematic errors, but possibly amplifies noise, and the weighted average introduces weight adjustment, but adopts fixed weight mostly, cannot adapt to spatial heterogeneity according to local conditions, and lacks independent quantization on each source error structure. Therefore, there is a need for a high-precision fusion method that can objectively evaluate errors, dynamically optimize weights, and promote multi-source T/ET consistency. Disclosure of Invention The application aims to provide a method, a system, equipment, a medium and a product for fusing a data set of transpiration and transpiration ratio based on triple collocation, the method effectively solves the problem of systematic deviation of multi-source T/ET data, remarkably improves estimation accuracy and enhances adaptability to regional heterogeneity. In order to achieve the above object, the present application provides the following. In a first aspect, the present application provides a method for fusing a data set of transpiration/transpiration ratios based on triple collocation, comprising: Three sets of transpiration T and transpiration ET data sets with different sources are obtained and preprocessed, so that three preprocessed T and ET data sets are obtained; For the three preprocessed T and ET data sets, respectively calculating error variances of T and ET in each data source by using a triple collocation method; distributing weights of T and ET in each data source according to the error variance of T and ET in each data source; Fusion calculation is carried out according to the weights of the T and the ET in each data source, and T and ET after fusion are obtained; And calculating the ratio of the T to the ET after fusion to obtain a T/ET data set after fusion. Optionally, the acquiring three sets of transpiration T and transpiration ET data sets from different sources and preprocessing are performed to obtain three preprocessed T and ET data sets, which specifically include: Three sets of T and ET data sets from different sources are obtained, space-time normalization processing is carried out on the data sets, the data sets are spatially unified into the same resolution, time scales are aligned, missing values are processed, a space mask is generated, only pixels with effective three source data are reserved, three preprocessed T and ET data sets are obtained, and the data of the three preprocessed data sources are respectively recorded as 、、And、、。 Optionally, for the three preprocessed T and ET data sets, a triple collocation method is used to calculate error variances of T and ET at each data source, which specifically includes: For data of three data sources after Tpre-processing 、、Using the formulaCalculating signal variance of TWhereinRepresentation ofAndIs a covariance of (2); Representation of AndIs a covariance of (2); Representation of AndIs a covariance of (2); Using the formula Calculating T at each data source、、Error variance of (2)、、Wherein、、Respectively represent、、Is a variance of (2); data for three data sources after ET preprocessing 、、Using the formulaCalcu