CN-122021328-A - Estimating ET0Weather type and wind force analysis coefficient optimization method

CN122021328ACN 122021328 ACN122021328 ACN 122021328ACN-122021328-A

Abstract

The invention discloses an ET 0 estimation method under the condition of public weather forecast data, which comprises the steps of obtaining daily weather observation data of years, obtaining weather forecast data of a plurality of weather stations distributed in a plurality of agricultural areas and the same period of the daily weather observation data, uniformly selecting weather stations in each agricultural area based on agricultural area division, collecting historical weather observation data of years, carrying out regional calibration on solar radiation parameters a s and b s by adopting a least square regression method to obtain solar radiation calculation parameters a s and b s month and annual calibration values of different agricultural areas, providing three public weather forecast information analysis strategies by depending on public resources, optimizing the weather types and the wind analysis coefficients, realizing accurate quantitative analysis of weather conditions to solar radiation quantity and wind power level to wind speed, and finishing ET 0 estimation with low cost and high precision.

Inventors

CAI JIABING
XUE XIN
WANG YIWEN
ZHANG BAOZHONG
WEI ZHENG
HAN CONGYING

Assignees

中国水利水电科学研究院

Dates

Publication Date: 20260512
Application Date: 20260205

Claims (10)

1. An optimization method for estimating the weather type and wind force analysis coefficient of ET 0 is used for realizing ET 0 estimation under the condition of public weather forecast data, and comprises the steps of obtaining daily weather observation data of years, and obtaining weather forecast data of a plurality of weather stations distributed in a plurality of agricultural areas and in synchronization with the daily weather observation data; The meteorological observation data comprise the highest air temperature, the lowest air temperature, relative humidity, solar radiation, sunshine hours and wind speed; the public weather forecast data comprises the highest air temperature, the lowest air temperature, weather conditions and wind directions, and is characterized in that the optimization method comprises the following steps: Under the condition of daily meteorological observation data, calculating based on a solar radiation actual measurement value R s to obtain a crop surface net radiation value R n in an ET 0 estimation formula, calculating by the ET 0 estimation formula to obtain an ET 0 estimation value under the condition of the observation data, and recording as ET 0 - PM ; Under the condition of public weather forecast data synchronous with daily meteorological observation data, calculating a crop surface net radiation value R n in an ET 0 estimation formula by a solar radiation calculation value R s ' determined by a public weather forecast data information analysis value, calculating an ET 0 estimation value under the condition of the public weather forecast data by an ET 0 estimation formula, and recording as an ET 0 - PWF ; And comparing ET 0 - PWF with ET 0 - PM , evaluating the estimation accuracy of ET 0 to obtain weather types and wind optimal analysis coefficients of a plurality of agricultural areas under the condition of public weather forecast data, and further realizing ET 0 high-accuracy estimation under the condition of public weather forecast data.
2. The optimization method according to claim 1, characterized in that, The calculation formula of the crop surface net radiation value R n is as follows: ; Wherein R n is the clean radiation of the surface of crops, R ns is the clean solar radiation, R nl is the clean output long-wave radiation; The determination of the net radiation value R n on the surface of the crop adopts the following two modes, and then an estimated value of ET 0 recorded as ET 0 - PWF and an estimated value of ET 0 recorded as ET 0 - PM are obtained: 1. Under the condition of meteorological observation data: ; ; ; ; Wherein R s is an actual measurement value of solar radiation, R a is zenith radiation, and R so is clear sky solar radiation; is the reflectivity or the canopy reflection coefficient; Is Stefan-Boltzmann constant, T max,k is the highest absolute temperature value within 24 hours, T min,k is the lowest absolute temperature value within 24 hours, e a is the actual water vapor pressure, a s is a regression constant, a s +b s is the part of the zenith radiation reaching the ground on a sunny day, N is the actual sunshine duration, N is the maximum possible sunshine duration or sunshine hours, G sc is the solar constant, d r is the relative distance between the days and the ground; is the inclination angle of the sun; is the local latitude; Is the sunset time angle; 2. Under the condition of public weather forecast data: ; ; ; Wherein R s ' is a solar radiation calculated value, R so is clear sky solar radiation, and R sc is total negative solar radiation; Analyzing coefficients for weather types; The ET 0 estimation formula is: ; Wherein R n is the net radiation value of the surface of crops, G is the soil heat flux density, T is the air temperature at the height of 2m, u 2 is the wind speed at the height of 2m, e s is saturated water vapor pressure, e a is actual water vapor pressure, and e s -e a is saturated water vapor pressure difference; Is the slope of the temperature-water vapor pressure curve; Is the thermometer constant; is the reflectivity or the canopy reflection coefficient; is Stefan-Boltzmann constant.
3. The optimization method according to claim 2, further comprising the step of regionalizing and calibrating parameters a s and b s by using the sunlight hours n and the actual measured value of solar radiation R s based on the daily observation data of a plurality of weather stations through the formula And (3) carrying out regional calibration on solar radiation parameters a s and b s of a plurality of agricultural areas by adopting a least square regression method of R s /R a and N/N to obtain a s and b s parameter values of the agricultural areas on the scale of each month and year.
4. The optimization method according to claim 1, wherein the weather type and wind optimum resolution coefficient obtaining method strategies include three kinds of method strategies, which are respectively: 1. Under the condition of fixed wind speed, adopting Monte Carlo iteration and differential evolution to realize quantitative analysis of weather conditions to solar radiation quantity, and optimizing weather type analysis coefficients; 2. On the premise of fixing weather type analysis coefficients, adopting Monte Carlo iteration and differential evolution to realize quantitative analysis of wind power grade to wind speed, and optimizing the wind power analysis coefficients; 3. under the condition of the double-factor cooperative change of the weather type and the wind power level, a particle swarm optimization algorithm and a self-adaptive particle swarm optimization algorithm are introduced to optimize the weather type and the wind power analysis coefficient.
5. The optimizing method according to claim 4, wherein the method step of adopting monte carlo iteration to realize quantitative analysis of weather conditions to solar radiation amount under the condition of fixed wind speed and optimizing the weather type analysis coefficient comprises the following steps: The method comprises the steps of firstly, fixing the wind speed to be the median value of a wind speed range with the height of 2m, determining an initial value range according to the empirical relation between the weather type and the weather type analysis coefficient, and setting the weather type analysis coefficient in a corresponding interval according to weather type classification; Secondly, randomly extracting a data set in a given corresponding interval according to uniform distribution, and carrying out 5000 Monte Carlo simulations to obtain 5000 weather type analysis coefficient candidate values; And thirdly, substituting each candidate value into an ET 0 estimation formula to calculate an ET 0 estimation value under the condition of public weather forecast data, comparing the ET 0 estimation value with an observed data ET 0 estimation value actually measured by a site, and selecting a value with the minimum average absolute error as an optimal analytic coefficient of the weather type.
6. The optimizing method according to claim 4, wherein the method step of optimizing the weather type analysis coefficient comprises the steps of: The method comprises the steps of firstly, fixing the wind speed to be the median value of a wind speed range with the height of 2m, and determining the global search interval of the weather type analysis coefficient according to the physical meaning of the weather type analysis coefficient; Initializing parameter populations in a determined global search interval range, generating candidate parameters through differential variation, recombining parameter fragments through cross operation, and substituting a generated test vector into an ET 0 estimation formula to calculate an objective function value, namely, reserving better parameters through selection operation, and repeating the evolution operation until an iteration termination condition is met; And thirdly, substituting the finally obtained weather type and the optimal analysis coefficient of the wind power into an ET 0 estimation formula to calculate an ET 0 value, comparing the ET 0 value with an estimated value of observed data ET 0 actually measured by a site, and verifying the fitting effect of the ET 0 value, wherein the value is a global optimal parameter under the corresponding weather type.
7. The optimizing method according to claim 4, wherein the method step of adopting monte carlo iteration to realize quantitative analysis of wind speed under wind power level under the condition of fixed weather type analysis coefficient, and optimizing the wind power analysis coefficient comprises the following steps: The method comprises the steps of firstly, fixing a weather type analysis coefficient as a median value of an initial interval of the weather type analysis coefficient, and determining the initial interval of the wind analysis coefficient based on a wind speed range at a position of 2m in the ground meteorological observation standard; secondly, randomly extracting a data set according to uniform distribution in a given wind force analysis coefficient initial interval range, and carrying out 5000 Monte Carlo simulation to obtain 5000 wind force analysis coefficient candidate values; and thirdly, substituting each candidate value into an ET 0 estimation formula to calculate an ET 0 estimation value under the condition of public weather forecast data, comparing the ET 0 estimation value with an observed data ET 0 estimation value actually measured by a website, and selecting a value with the minimum average absolute error as an optimal analytic coefficient of the wind power grade.
8. The optimizing method according to claim 4, wherein the method step of optimizing the wind force analysis coefficient by quantitatively analyzing the wind force level to the wind speed using differential evolution under the condition of a fixed weather type analysis coefficient comprises: The method comprises the steps of firstly, fixing a weather type analysis coefficient as a median value of an initial interval of the weather type analysis coefficient, and determining a global search interval of the wind analysis coefficient based on a wind speed range at a height of 2m given in ground meteorological observation standards under different wind power grade conditions according to the physical meaning of the wind analysis coefficient so as to cover a reasonable physical interval; Initializing parameter population in the determined search interval range, generating candidate parameters through differential variation, recombining parameter fragments through cross operation, and substituting the generated test vector into an ET 0 estimation formula to calculate an objective function value, namely, reserving better parameters through selection operation, and repeating the evolution operation until the iteration termination condition is met; And thirdly, substituting the obtained optimal wind force analysis coefficient into an ET 0 estimation formula, and comparing the optimal wind force analysis coefficient with an observed data ET 0 estimated value actually measured by a site to verify the fitting effect, wherein the value is a global optimal parameter under the corresponding wind force grade.
9. The optimization method according to claim 4, wherein the step of the particle swarm optimization algorithm comprises: determining a double-parameter search range according to the weather type, the weather type analysis coefficient, the wind power level and the experience constraint relation of the wind power analysis coefficient, wherein the value interval of the weather type analysis coefficient is an interval range corresponding to 10 weather types, and the value interval of the wind power analysis coefficient is a 2m high wind speed range given in the ground weather observation Specification; Secondly, initializing a particle swarm in the interval value and the wind speed range at the height of 2m, setting fixed inertia weight and learning factors, and updating the speed and the position of the particles by combining the individual historical optimal position and the population global optimal position; And thirdly, after iteration is finished, selecting the parameter combination with the minimum fitness MAE in the population as the optimal weather type and wind force analysis coefficient under the condition of weather type and wind force grade.
10. The optimization method according to claim 4, wherein the step of the adaptive particle swarm optimization algorithm under the condition of the two-factor cooperative variation of the weather type and the wind level comprises: determining a double-parameter search range according to the weather type, the weather type analysis coefficient and the physical constraint relation between the wind power level and the wind power analysis coefficient, wherein the value interval of the weather type analysis coefficient is an interval range corresponding to 10 weather types, and the value interval of the wind power analysis coefficient is a 2m high wind speed range given in the ground weather observation Specification; Initializing a particle swarm in a determined interval value and a wind speed range at a height of 2m, adaptively adjusting key parameters in an iterative process, wherein inertia weight linearly decreases from 0.9 to 0.4, learning factors are dynamically corrected according to particle fitness, and particles are updated at a speed and a position by combining 'individual optimal + global optimal'; And thirdly, after iteration is finished, selecting a parameter combination with the minimum fitness MAE in the population, substituting the parameter combination into an ET 0 estimation formula, and carrying out fitting inspection with an observed data ET 0 estimated value actually measured by a site to serve as an optimal weather type and a wind power analysis coefficient under the condition of weather type and wind power level.

Description

Optimization method for estimating weather type and wind power analysis coefficient of ET 0 Technical Field The invention relates to an optimization method for estimating the weather type and the wind power analysis coefficient of ET 0, which is based on public weather forecast to estimate the weather type and the wind power analysis coefficient of ET 0. Background Background The weather forecast information is used for estimating ET 0, so that the water demand condition of farmland crops can be estimated in real time, and the irrigation strategy can be adjusted in time. Weather forecast information is mainly divided into public weather forecast and numerical weather forecast. The public weather forecast mainly utilizes historical data, a statistical model and the experience of a weather student to predict the future weather change of a site, and generally comprises information such as the highest air temperature, the lowest air temperature, weather conditions, wind power and the like. The numerical weather forecast is a method for predicting future weather changes by solving a weather motion equation set through a mathematical physical model and a computer technology, and generally comprises the steps of obtaining weather forecast data of any point on a grid through interpolation calculation, wherein the data format of the weather forecast is a grid format of related weather variables such as temperature, water vapor pressure, solar radiation and the like. Numerical weather forecast has higher spatial resolution and more weather parameters, but the deviation of initial values, simplification of equations or low predictability of certain factors can cause weather forecast failure, so that the ET 0 forecast is performed based on the weather forecast, and the forecast capability of a single numerical weather forecast model is improved by checking variables and stations or adopting a method of forecasting by using different forecasting systems together. Meanwhile, the numerical weather forecast has the problem of inconsistent forecast, namely the phenomenon of larger difference between two forecast results. Therefore, the pretreatment of the numerical weather forecast is complex, and certain difficulty is brought to wide use. Public weather forecast can be obtained freely on common media, and forecast information can be used for ET 0 calculation through digital analysis. Cai Jiabing ET al (2005) propose an analytical method to convert public weather forecast into weather variables needed to calculate daily ET 0 using Penman-montith (PM) equation (FAO-56). The method is characterized in that weather types are classified into five types of sunny, cloudy, overcast gust and continuous overcast, 5 weather types are assumed to be evenly distributed between 0 radiation and sunny radiation Rso, corresponding sunshine hours and solar radiation values are obtained through analysis, and wind power level is fixed to be the median value of a 2m high wind speed range given by ground weather observation Specification. Through the processing, the weather forecast information is converted into a complete PM equation input item, and the feasibility of the method is verified after the weather forecast information is compared with the measured weather station data. The analysis method simplifies the calculation flow, saves the cost and is convenient for popularization and application. However, when the weather condition is analyzed into the sunshine hours or the solar radiation quantity, a fixed and empirical corresponding relation is adopted, so that the regional adaptability is poor. The current shortcomings of estimating ET 0 using public weather forecast are mainly manifested in: 1) When the weather condition is analyzed into solar radiation quantity and the wind power grade is analyzed into wind speed in public weather forecast for irrigation management, certain deviation exists in different agricultural areas, so that ET 0 estimation accuracy is reduced. 2) When the solar radiation Rs is calculated, the key parameters as and bs in the Angstrom formula are not regional rated, and the estimated value deviation can be generated by adopting the default value to calculate under different climates and geographic conditions, so that the ET 0 precision is reduced. Disclosure of Invention The invention aims to solve the problems of insufficient parameter universality and low ET 0 estimation accuracy in the existing public weather forecast analysis process, and provides an optimization method for estimating the weather type and wind force analysis coefficient of ET 0 so as to realize accurate estimation of ET 0 in different agricultural areas under the condition of public weather forecast data. In order to achieve the above object, the present invention is provided with: An optimization method for estimating the weather type and wind force analysis coefficient of ET 0 is used for realizing the accurate estimation of ET 0 under the co