Cotton high-temperature meteorological disaster prediction method

Abstract

The invention discloses a cotton high-temperature meteorological disaster prediction method, and relates to the technical field of cotton high-temperature meteorological disaster prediction, and the method comprises the steps: obtaining the high-temperature disaster information of cotton in different growth stages and the planting environment information when the high-temperature disaster occurs; taking the high-temperature disaster information of each growth stage of the cotton as a response variable, taking the planting environment information when the high-temperature disaster occurs as an explanatory variable, calculating a correlation coefficient between each explanatory variable and the response variable, and constructing a linear regression model based on the screened explanatory variables and response variables, the high-temperature disaster change of the cotton in the corresponding growth stage is predicted. According to the method, quantitative analysis on high-temperature disaster influence is realized, and the specific influence degree of high temperature on each growth stage of cotton is determined.

Description

Technical Field

[0001] The present invention relates to the technical field of cotton high - temperature meteorological disaster prediction, and specifically to a method for predicting cotton high - temperature meteorological disasters. Background Technique

[0002] Xinjiang is an important cotton - growing area in China. Due to its long sunshine hours and little rainfall, it is very suitable for cotton growth. Cotton in Xinjiang is generally sown from late April to mid - May and enters the picking period from September to October. July to August is a stage with frequent high - temperature weather in Xinjiang, which just covers multiple growth and development stages of cotton. The impact of high temperature on cotton is complex, including both the impact on the external environment (such as soil moisture content and planting density) and the direct impact on the growth and development status of cotton itself (such as germination rate, seedling chlorosis rate, flower organ deformation rate, and boll maturity).

[0003] However, current methods are mostly based on the assessment of overall weather conditions, that is, simply predicting and evaluating high - temperature disasters through meteorological data, and cannot provide phased impact data based on specific high - temperature thresholds. They fail to conduct refined analysis according to the characteristics of each growth stage of cotton (such as germination stage, seedling stage, flowering stage, and boll - setting stage). Growers cannot clearly identify the actual threat level of different degrees of high temperature to cotton growth, and it is difficult to establish a targeted early - warning system and disaster response mechanism. In addition, high - temperature disasters are usually accompanied by dynamic changes in the environment, such as a decrease in soil moisture. The superimposed effects of these factors and high temperature will exacerbate the impact on cotton. Existing technologies fail to incorporate these dynamic environmental factors into the evaluation scope, resulting in a large deviation in the comprehensive analysis of disaster impacts. Summary of the Invention

[0004] I) Technical Problems to be Solved

[0005] To achieve the above object, the present invention provides the following technical solution: A method for predicting cotton high - temperature meteorological disasters, including:

[0006] Setting high - temperature thresholds corresponding to different growth stages of cotton. During different growth stages of cotton, when the air temperature is higher than the corresponding high - temperature threshold, obtaining high - temperature disaster information of cotton and planting environment information at the time of high - temperature disaster occurrence; wherein, the high - temperature disaster information characterizes the corresponding characteristics of cotton affected by high - temperature disasters at different growth stages, and the planting environment information includes temperature information, soil depth, and planting density;

[0007] Using the high - temperature disaster information of each growth stage of cotton as the response variable and the planting environment information at the time of high - temperature disaster occurrence as the explanatory variable; wherein, using the temperature information as the key explanatory variable and the soil depth and planting density as other explanatory variables;

[0008] At each growth stage of cotton, calculate the correlation coefficient between each of the explanatory variables and the response variable, and select the explanatory variables with the absolute value of the correlation coefficient greater than a set threshold as the explanatory variables that have a significant impact on the response variable;

[0009] At each growth stage of cotton, construct a linear regression model based on the selected explanatory variables and the response variable;

[0010] According to the linear regression model constructed for each growth stage, input the actual planting environment information to predict the specific impact of high-temperature disasters on cotton during the corresponding growth stage.

[0011] Furthermore, take one cotton planting field as one planting unit, and the cotton planting varieties within each planting unit are the same. Among multiple planting units, obtain the high-temperature disaster information of the cotton at different growth stages and the planting environment information when high-temperature disasters occur.

[0012] Furthermore, obtain the high-temperature information of the cotton when high-temperature disasters occur at different growth stages through the data platform of the National Meteorological Bureau. The high-temperature information includes daily maximum temperature data.

[0013] Furthermore, for each growth stage of cotton, standardize the response variables and explanatory variables obtained from multiple planting units to convert response variables and explanatory variables of different scales into values of the same scale; the response variables corresponding to different growth stages are:

[0014] At the germination stage, obtain the germination rate of cotton as the response variable;

[0015] At the seedling stage, obtain the yellowing rate of cotton seedlings as the response variable;

[0016] At the flowering stage, obtain the deformation rate of cotton flower organs as the response variable;

[0017] At the boll-setting stage, obtain the damaged rate of cotton bolls as the response variable.

[0018] Furthermore, at each growth stage of cotton, calculate the correlation coefficient between each of the explanatory variables and the response variable. Specifically, use the Pearson correlation coefficient formula. After converting response variables and explanatory variables of different scales into values of the same scale at each growth stage of cotton, calculate the correlation coefficient r between each explanatory variable and the response variable respectively:

[0019]

[0020] where X i is the standardized explanatory variable, Y i is the standardized response variable, and is the mean value;

[0021] Select the explanatory variables whose absolute value of the correlation coefficient r is greater than the threshold as the explanatory variables that have a significant impact on the response variable.

[0022] Furthermore, after performing Pearson correlation analysis, a linear regression model for each growth stage of cotton is constructed to analyze the relationship between the response variable and the selected explanatory variables;

[0023] The formula of the linear regression model is:

[0024] Y = β 0 + β 1 X 1 + … + β i X i + ∈

[0025] Where: Y is the standardized response variable;

[0026] X i is the standardized explanatory variable;

[0027] β 0 is the constant term;

[0028] β i is the regression coefficient of the standardized explanatory variable;

[0029] ∈ is the error term.

[0030] Furthermore, the regression coefficient reflects the influence degree of different explanatory variables on the response variable, and the regression coefficients of each explanatory variable are calculated by the least squares method; specifically:

[0031] Construct a response variable matrix B and an explanatory variable matrix A according to the standardized response variable and explanatory variables;

[0032] Use the least squares formula to calculate the regression coefficient:

[0033] β = (A T A) -1 A T B

[0034] Where: A T is the transpose matrix of the explanatory variable matrix A;

[0035] (A T A) -1 is the inverse matrix of A T A.

[0036] Further, the linear regression model constructed according to each growth stage predicts the impact of high-temperature disasters on cotton during the corresponding growth stage, specifically as follows:

[0037] In the germination stage, taking the germination rate as the response variable, a linear regression model of the explanatory variables that have a significant impact on the germination rate after screening is constructed, and according to this model, the impact of high-temperature disasters on cotton germination during the germination stage is predicted;

[0038] In the seedling stage, taking the seedling chlorosis rate as the response variable, a linear regression model of the explanatory variables that have a significant impact on the seedling chlorosis rate after screening is constructed, and according to this model, the impact of high-temperature disasters on cotton seedling development during the cotton seedling stage is predicted;

[0039] In the flowering stage, taking the flower organ deformation rate as the response variable, a linear regression model of the explanatory variables that have a significant impact on the flower organ deformation rate after screening is constructed, and according to this model, the impact of high-temperature disasters on cotton flower organ development during the cotton flowering stage is predicted;

[0040] In the boll-setting stage, taking the boll damage rate as the response variable, a linear regression model of the explanatory variables that have a significant impact on the boll damage rate after screening is constructed, and according to this model, the impact of high-temperature disasters on cotton boll fruit maturity during the cotton boll-setting stage is predicted.

[0041] III) Beneficial effects:

[0042] Compared with the prior art, the invention has the following beneficial effects:

[0043] By obtaining cotton at each growth stage, extracting specific high-temperature disaster indicators as response variables respectively, using the temperature information and planting environment information during high-temperature disasters as explanatory variables, screening out the explanatory variables that have a significant impact on the response variables through correlation analysis, constructing a linear regression model, and combining the high-temperature information, planting environment information with the high-temperature disaster indicators at different growth stages of cotton, the quantitative analysis of the impact of high-temperature disasters is realized, and the specific impact degree of high temperature on each growth stage of cotton is clarified. BRIEF DESCRIPTION OF THE DRAWINGS

[0044] Figure 1 It is a flowchart of the cotton high-temperature meteorological disaster prediction method provided by the embodiment of the present invention;

[0045] Figure 2 It is a schematic flow chart of screening the explanatory variables that have a significant impact on the response variable in the cotton high-temperature meteorological disaster prediction method provided by the embodiment of the present invention;

[0046] Figure 3 It is a schematic technical route diagram of the cotton high-temperature meteorological disaster prediction method provided by the embodiment of the present invention. DETAILED DESCRIPTION OF THE INVENTION

[0047] The following will clearly and completely describe the technical solutions in the embodiments of the present invention in conjunction with the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without making creative efforts shall fall within the protection scope of the present invention.

[0048] In the description of the present invention, it should be understood that the orientation or positional relationships indicated by the terms "longitudinal", "transverse", "length", "width", "thickness", "upper", "lower", "front", "rear", "left", "right", "vertical", "horizontal", "top", "bottom", "inner", "outer", etc. are based on the orientation or positional relationships shown in the accompanying drawings, and are only for the convenience of describing the present invention and simplifying the description, rather than indicating or implying that the device or element referred to must have a specific orientation, be constructed and operated in a specific orientation, and thus should not be construed as a limitation to the present invention.

[0049] In addition, terms such as "first" and "second" are only used for distinguishing descriptions and should not be construed as indicating or implying relative importance.

[0050] It should be noted that, without conflict, the features in the embodiments of the present invention can be combined with each other.

[0051] High temperature not only affects the external environment of cotton planting, but also has a significant impact on the growth and development of cotton itself at different growth stages. Taking Xinjiang as an example, the climate in Xinjiang belongs to a typical arid continental climate, with large temperature differences between day and night, long sunshine hours, and little rainfall. This climate characteristic is very suitable for cotton planting. In particular, sufficient sunshine helps the growth of cotton fibers.

[0052] Although the climate in Xinjiang is suitable for cotton planting, in summer when the temperature is high, meteorological indicators judge that a high temperature greater than 35°C and lasting for 3 days is a mild high temperature heat damage, lasting for 5 days is a moderate high temperature heat damage, and lasting for 7 days or more is a severe high temperature heat damage. In the case of high temperature exceeding 35°C, high temperature will have an adverse impact on different growth stages of cotton. For example:

[0053] In the germination stage of cotton, high temperature may cause uneven seed germination, and even the situation where seeds cannot grow and develop, resulting in a decrease in the germination rate.

[0054] In the seedling stage of cotton, high temperature causes excessive water loss in the leaves, resulting in yellowing or withering, and an increase in the yellowing rate of seedlings.

[0055] In the flowering stage of cotton, high temperature causes the petals to deform or develop incompletely, affecting pollination, pollination failure, and flower abscission, thereby resulting in an increase in the deformation rate of flower organs.

[0056] During the boll-setting stage of cotton, high temperature affects the normal development of bolls, resulting in short, thin fibers or reduced quality, thereby increasing the damage rate of bolls.

[0057] Combined with Figures 1 to 3 As shown, the embodiment of the present invention proposes a method for predicting high-temperature meteorological disasters in cotton, which is used to construct a prediction model according to different high-temperature disaster characteristics and environmental factors at each growth stage of cotton, so as to accurately intervene and reduce the harm of high temperature to different growth stages of cotton. The following method steps are specifically implemented.

[0058] Step S10: Set high-temperature thresholds corresponding to different growth stages of cotton. When the temperature is higher than the corresponding high-temperature threshold during different growth stages of cotton, obtain the high-temperature disaster information of cotton and the planting environment information when the high-temperature disaster occurs; among them, the high-temperature disaster information characterizes the corresponding characteristics of cotton affected by high-temperature disasters at different growth stages, and the planting environment information includes temperature information, soil depth, and planting density.

[0059] Combined with the above, it can be understood that different disaster characteristics are generated by high-temperature disasters in different growth stages of cotton. For example, when affected by high-temperature disasters, the germination rate of cotton seeds decreases during the germination stage, the yellowing rate of cotton seedlings increases during the seedling stage, the deformation rate of cotton flower organs increases during the flowering stage, and the damage rate of cotton bolls increases during the boll-setting stage. Therefore, the germination rate, the yellowing rate of seedlings, the deformation rate of flower organs, and the damage rate of bolls are respectively used as the high-temperature disaster information for each growth stage. Because the planting densities in different cotton planting areas are different and the planting density will also have a certain impact on the growth and development of cotton, the soil humidity and planting density are used as the planting environment information.

[0060] It should be noted that because the threshold definitions of high-temperature disasters vary from place to place and the high-temperature weather values affecting the development status of cotton in each growth stage are also the same, the high-temperature information obtained when the high-temperature disaster occurs is a variable value, which is defined according to the weather conditions in different regions. Therefore, in the embodiments provided by the present invention, no necessary limitation is imposed on the temperature threshold of the high-temperature information.

[0061] In summary, in some feasible embodiments provided by the present invention, during the data acquisition process, cotton fields with the same type of cotton planting are selected, and one cotton field is used as a planting unit. From multiple planting units, the high-temperature disaster information of cotton in different growth stages is collected respectively. Specifically:

[0062] During the germination stage: Collect the germination rate of cotton;

[0063] During the seedling stage: Collect the yellowing rate of cotton seedlings;

[0064] During the flowering stage: collect the deformation rate of cotton flower organs;

[0065] During the boll-setting stage: collect the damage rate of cotton bolls.

[0066] Meanwhile, collect the soil moisture and cotton planting density at the time of high temperature occurrence.

[0067] Through the data platform of the National Meteorological Bureau, obtain the daily maximum temperature data of cotton during high temperature disasters at different growth stages. These meteorological data can be used to determine the high temperature impact in different time periods and provide data support for key explanatory variables in subsequent models.

[0068] After data acquisition is completed, perform step S20: Use the high temperature disaster information at each growth stage of cotton as the response variable, and use the planting environment information at the time of high temperature disaster as the explanatory variable; among them, use temperature information as the key explanatory variable, and use soil depth and planting density as other explanatory variables.

[0069] It can be understood in combination with the above that specifically refer to Figure 3 , during the germination stage of cotton, use the collected germination rate as the response variable; during the seedling stage of cotton, use the collected seedling chlorosis rate as the corresponding variable; during the flowering stage of cotton, use the collected deformation rate of flower organs as the response variable; during the boll-setting stage of cotton, use the collected damage rate of cotton bolls as the response variable.

[0070] Use the collected high temperature information, soil moisture, and planting density as explanatory variables. Among them, because high temperature has the greatest impact on the growth and development of cotton, high temperature information is used as the key explanatory variable, and soil moisture and planting density are used as other explanatory variables.

[0071] Then perform step S30: At each growth stage of cotton, calculate the correlation coefficient between each explanatory variable and the response variable, and screen out the explanatory variables with the absolute value of the correlation coefficient greater than the set threshold as the explanatory variables that have a significant impact on the response variable.

[0072] To sum up, due to the different scales of explanatory variables such as high temperature, soil moisture, and planting density, standardize these explanatory variables first to make them have the same scale, which can avoid biases caused by different variable units. Then, perform correlation analysis to ensure that the standardized data can be fairly compared in terms of correlation, judge the collinearity problem between variables, and screen out redundant variables.

[0073] It can be understood that when the dimensional differences of different variables are large, for example, temperature is in °C, soil moisture is in percentage, and planting density is in the number of plants per unit area. Standardization can convert the scales of all variables to the same range (such as mean 0, standard deviation 1), ensuring that the impacts of different variables can be fairly compared.

[0074] It should be noted that the standardization operation of data should be carried out before the correlation analysis of variables. This is because the standardized different variable values help to reduce the deviation between different explanatory variables and make the correlation analysis results more reliable. If not standardized, the numerical values of some variables are relatively large, which may play a dominant role in calculating the correlation, thus affecting the correct evaluation of other variables. For example, the high temperature value may be between 35°C and 40°C, while the planting density value may be in the range of 0 to 1000. If this numerical difference is not processed, it will cause misleading in the correlation analysis.

[0075] In some feasible embodiments provided by the present invention, the response variables and explanatory variables at each growth stage of the collected cotton are subjected to Z-score standardization processing, that is, the data is transformed into the form of "standard normal distribution", that is, a distribution with a mean of 0 and a standard deviation of 1. After Z-score standardization, the center of the data moves to 0 (i.e., the mean is 0), and is scaled according to the standard deviation, that is, the response variables and explanatory variables of different scales are transformed into values of the same scale.

[0076] Specifically, use Z-score standardization:

[0077]

[0078] Among them, Z is the value after standardization, x i is the original value, that is, the response variables and explanatory variables of different scales, μ is the mean of the variable, and σ is the standard deviation of the variable.

[0079] Now, taking the data collected from the seedling growth stage of cotton as an example for detailed explanation. At the seedling stage:

[0080] Response variable (Y 1 ): Seedling chlorosis rate (%), which represents the proportion of chlorotic seedlings in the total seedlings affected by high temperature disasters;

[0081] Key explanatory variable (X 1 ): High temperature (daily maximum temperature, unit: °C). Now set the high temperature threshold to 35°C, and record the number of high temperature days and the corresponding temperature when the daily maximum temperature exceeds this threshold.

[0082] Other explanatory variables (X 2 and X 3 ): Soil humidity (%, currently set as the water content of the soil at 10 cm below the ground surface), planting density (plants per square meter, representing the number of cotton plants planted per square meter).

[0083] Data collected from five planting units of the same cotton variety but different planting locations are shown in the following table:

[0084]

[0085]

[0086] Standardize each variable so that its mean is 0 and standard deviation is 1 to eliminate the difference in dimensions, and obtain the standardized data table:

[0087]

[0088] After standardizing the response variable and explanatory variables, perform a correlation analysis. The correlation analysis is used to determine the strength of the relationship between each explanatory variable (high temperature, soil humidity, planting density) and the response variable. Screening out the variables that have a significant impact on the response variable and ensuring that only the explanatory variables with strong correlations are included in the regression model helps to improve the accuracy and simplicity of the model.

[0089] Taking the data collected at the seedling stage of the above-mentioned cotton as an example, analyze the correlations between high temperature, soil humidity, and planting density and the seedling chlorosis rate.

[0090] In some feasible embodiments of the invention, use the Pearson correlation coefficient formula to calculate the correlation coefficient r between each explanatory variable and the response variable;

[0091]

[0092] where X i and Y i are the explanatory variable and the response variable respectively, and are their means.

[0093] According to the standardized data table obtained above, using the Pearson correlation coefficient calculation gives:

[0094] The correlation coefficient between the average high temperature and the seedling chlorosis rate is 0.45;

[0095] The correlation coefficient between the number of high-temperature days and the seedling chlorosis rate is 0.82;

[0096] The correlation coefficient between the soil humidity and the seedling chlorosis rate is -0.55;

[0097] The correlation coefficient between the planting density and the seedling chlorosis rate is 0.25;

[0098] Now screen out the explanatory variables with the absolute value of the correlation coefficient r greater than 0.4 as the explanatory variables that have a significant impact on the response variable.

[0099] Thus, it can be concluded that:

[0100] The absolute values of the correlation coefficients between the number of high-temperature days and the average high temperature and the seedling yellowing rate are both greater than 0.4, indicating that high-temperature information, as a key explanatory variable, the number of high-temperature days and the average high temperature are the main influencing factors of the yellowing rate, and the correlation is extremely high.

[0101] The correlation coefficient between soil humidity and the yellowing rate is -0.55, and its absolute value is greater than 0.4, indicating that a decrease in humidity will lead to an increase in the yellowing rate, and the correlation is relatively high.

[0102] The correlation coefficient between planting density and the yellowing rate is 0.25, and its absolute value is less than 0.4, indicating that an increase in planting density also has a positive impact on the yellowing rate, but the degree of influence is small and not considered relevant.

[0103] In summary, the screening results: the number of high-temperature days, the average high temperature, and soil humidity are significant explanatory variables.

[0104] It can be understood that setting a correlation coefficient threshold is to select variables with a linear relationship above medium with the response variable. If the correlation coefficient is too low, it means that this explanatory variable contributes less to the response variable, may form multicollinearity with other variables, and affect the stability and interpretability of the regression model. Therefore, selecting explanatory variables with a higher correlation coefficient can reduce the impact of multicollinearity, improve the reliability of the model, and too many explanatory variables will not only increase the complexity of the model, but may also lead to the phenomenon of "overfitting" and affect the prediction performance of the model. The common Pearson correlation coefficient interpretations are as follows:

[0105] 0.1 - 0.3: Weak correlation;

[0106] 0.3 - 0.5: Medium correlation;

[0107] 0.5 - 1.0: Strong correlation.

[0108] In some embodiments, setting the correlation threshold to 0.4 means retaining explanatory variables with at least a medium degree of correlation with the response variable, thereby improving the accuracy and stability of the model. For variables with an absolute value of the correlation coefficient less than 0.4, their influence on the response variable is small, so they are not selected into the model.

[0109] After standardization processing and screening the explanatory variables that have a significant impact on the response variable, step S40 is implemented: in each growth stage of cotton, based on the screened explanatory variables and the response variable, a linear regression model is constructed.

[0110] Specifically, in some feasible embodiments of the present application, after completing the Pearson correlation analysis, a linear regression model is then used to further calculate the weight coefficients of the explanatory variables (high temperature, soil humidity, planting density) for the response variable. It should be noted that the explanatory variables for constructing the model are the selected explanatory variables that have a significant impact on the response variable, so as to construct a more accurate prediction model.

[0111] Linear regression equations applicable to each growth stage:

[0112] Y = β 0 + β 1 X 1 + β 2 X 2 + β 3 X 3 + ∈

[0113] Where: Y is the response variable; X 1 、X 2 and X 3 are the standardized explanatory variables (number of high-temperature days, average high temperature, soil humidity) respectively; β 0 is the constant term (intercept); β 1 、β 2 and β 3 are the regression coefficients (weight coefficients) of each standardized explanatory variable; ∈ is the error term.

[0114] Among them, in some feasible embodiments of the present invention, the least squares method is used to calculate the regression coefficients β 1 、β 2 and β 3 .

[0115] Still taking the response variable data and explanatory variable data collected in the cotton seedling stage as an example above:

[0116] β 0 is the intercept, indicating the expected value of the seedling chlorosis rate when all explanatory variables are 0;

[0117] β 1 represents the amount of influence on the seedling chlorosis rate for each additional day of high-temperature days;

[0118] β 2 represents the amount of influence on the seedling chlorosis rate for each additional 1°C of average high temperature;

[0119] β 3 represents the amount of influence on the seedling chlorosis rate for each additional 1% of soil humidity.

[0120] First, construct a response variable matrix B and an explanatory variable matrix A based on the standardized response variable and explanatory variables:

[0121]

[0122] Among them: the first column is the intercept term, the second column is the number of high-temperature days after standardization, the third column is the average high temperature after standardization, and the fourth column is the soil humidity after standardization.

[0123]

[0124] Calculate the regression coefficient using the least squares formula:

[0125] β = (A T A) -1 A T B

[0126] Among them: A T is the transpose matrix of the explanatory variable matrix A;

[0127] (A T A) -1 is the inverse matrix of A T A.

[0128] Obtain the regression coefficient β through matrix multiplication:

[0129]

[0130] Obtain the regression equation as:

[0131] Y = 1.45X 1 - 0.96X 2 - 0.32X 3

[0132] Among them:

[0133] β 1 = 1.45: indicating that the positive effect of the number of high-temperature days on the seedling chlorosis rate is the most significant;

[0134] β 2 = 0.96: indicating that the positive effect of high temperature on the seedling chlorosis rate is the second;

[0135] β 3 = - 0.32: indicating the negative effect of soil humidity on the seedling chlorosis rate, indicating that high soil humidity can reduce the chlorosis rate.

[0136] Through this linear regression equation, the change of the etiolation rate in the seedling stage can be predicted based on different explanatory variables. For example, in the seedling stage, the number of high-temperature days, average high temperature, and soil humidity are obtained from the actual situation. After standardizing these data and substituting them into the regression equation to obtain the predicted value Y, the predicted value Y is then converted back to the actual value, that is, an inverse standardization operation is performed on the predicted value Y to restore the prediction result to a value with practical significance for subsequent intuitive interpretation and application.

[0137] In summary, it can be understood that in the germination stage of cotton, the germination rate collected is the response variable, and the significant explanatory variables are selected through correlation analysis. Then, a linear regression model that conforms to the law of the impact of high-temperature disasters in this stage is constructed to predict the impact of high-temperature disasters on cotton germination in the germination stage.

[0138] In the seedling stage of cotton, the etiolation rate of seedlings collected is the response variable, and the significant explanatory variables are selected through correlation analysis. Then, a linear regression model that conforms to the law of the impact of high-temperature disasters in this stage is constructed, and the model is used to predict the impact of high-temperature disasters on the development of cotton seedlings in the seedling stage.

[0139] In the flowering stage of cotton, the deformation rate of flower organs collected is the response variable, and the significant explanatory variables are selected through correlation analysis. Then, a linear regression model that conforms to the law of the impact of high-temperature disasters in this stage is constructed, and the model is used to predict the impact of high-temperature disasters on the development of cotton flower organs in the flowering stage.

[0140] In the boll-setting stage of cotton, the damage rate of cotton bolls collected is the response variable, and the significant explanatory variables are selected through correlation analysis. Then, a linear regression model that conforms to the law of the impact of high-temperature disasters in this stage is constructed, and the model is used to predict the impact of high-temperature disasters on the ripening of cotton bolls in the boll-setting stage.

[0141] Through the linear regression models of each growth stage, the specific impact of high-temperature disasters on cotton growth in each growth stage is obtained, and effective management and intervention are carried out according to the results predicted by the models. This will help improve the yield and quality of cotton and reduce potential losses caused by high temperature.

[0142] The above is only a preferred embodiment of the present invention and is not intended to limit the present invention. The scope of patent protection of the present invention is subject to the claims. All equivalent structural changes made by using the description and drawings of the present invention shall be equally included in the protection scope of the present invention.

Claims

1. A method for predicting high temperature meteorological disasters for cotton, characterized in that: include: Setting high temperature thresholds corresponding to different growth stages of cotton, and obtaining high temperature disaster information of cotton and planting environment information when high temperature disasters occur when the temperature is higher than the corresponding high temperature thresholds at different growth stages of cotton; wherein the high temperature disaster information represents the corresponding characteristics of cotton affected by high temperature disasters at different growth stages, and the planting environment information includes temperature information, soil depth and planting density; The high temperature disaster information at each growth stage of cotton is used as the response variable, and the planting environment information when the high temperature disaster occurs is used as the explanatory variable; wherein the temperature information is used as the key explanatory variable, and the soil depth and planting density are used as other explanatory variables; At each growth stage of cotton, the correlation coefficient between each of the explanatory variables and the response variable is calculated, and the explanatory variables whose absolute value of the correlation coefficient is greater than a set threshold are screened out as the explanatory variables that have a significant impact on the response variable; At each growth stage of cotton, a linear regression model is constructed based on the screened explanatory variables and response variables; According to the linear regression model constructed in each growth stage, the actual planting environment information is input to predict the specific impact of high temperature disasters on cotton in the corresponding growth stage.

2. A cotton high temperature meteorological disaster prediction method according to claim 1, characterized in that: A cotton planting field is regarded as a planting unit, and the cotton planting varieties in each planting unit are the same. In multiple planting units, high temperature disaster information of the cotton at different growth stages and planting environment information when high temperature disaster occurs are obtained respectively.

3. A cotton high temperature meteorological disaster prediction method according to claim 1, characterized in that: High temperature information when high temperature disasters occur in different growth stages of the cotton is obtained through the National Meteorological Administration data platform, and the high temperature information includes daily maximum temperature data.

4. A cotton high temperature meteorological disaster prediction method according to claim 1, characterized in that: For each growth stage of cotton, the response variables and explanatory variables obtained from the plurality of planting units are standardized, and the response variables and explanatory variables of different scales are converted into values ​​of the same scale; the response variables corresponding to different growth stages are: At the germination stage, the germination rate of cotton was obtained as the response variable; At the seedling stage, the yellowing rate of cotton seedlings was obtained as the response variable; At the flowering stage, the floral deformation rate of cotton was obtained as the response variable; At the boll-setting stage, the boll damage rate of cotton was obtained as the response variable.

5. A cotton high temperature meteorological disaster prediction method according to claim 4, characterized in that: At each growth stage of cotton, the correlation coefficient between each explanatory variable and the response variable is calculated. Specifically, the Pearson correlation coefficient formula is used to convert the response variables and explanatory variables of different scales into values ​​of the same scale at each growth stage of cotton, and then the correlation coefficient r between each explanatory variable and the response variable is calculated respectively: Among them, X i is the standardized explanatory variable, Y i is the standardized response variable, and is the mean; Explanatory variables whose absolute values ​​of correlation coefficients r are greater than the threshold are selected as explanatory variables that have a significant impact on the response variable.

6. A method for predicting high temperature meteorological disasters for cotton according to claim 5, characterized in that: After Pearson correlation analysis, a linear regression model for each growth stage of cotton was constructed to analyze the relationship between the response variable and the screened explanatory variables; The linear regression model formula is: Y=β0+β1X1+…+β i X i +∈ Where: Y is the standardized response variable; X i is the standardized explanatory variable; β0 is a constant term; β i The regression coefficients of the standardized explanatory variables; ∈ is the error term.

7. A method for predicting high temperature weather disasters for cotton according to claim 6, characterized in that: The regression coefficient reflects the influence of different explanatory variables on the response variable, and the regression coefficient of each explanatory variable is calculated by the least squares method; specifically: Constructing a response variable matrix B and an explanatory variable matrix A according to the standardized response variables and explanatory variables; The regression coefficients are calculated using the least squares formula: β=(A T A) -1 A T B Among them: A T is the transposed matrix of the explanatory variable matrix A; (A T A) -1 Yes A T The inverse matrix of A.

8. A method for predicting high temperature meteorological disasters for cotton according to claim 7, characterized in that: The linear regression model constructed according to each growth stage predicts the impact of high temperature disasters on cotton in the corresponding growth stage, specifically: In the germination stage, taking the germination rate as the response variable, a linear regression model with the selected explanatory variables that significantly affect the germination rate was constructed, and the impact of high temperature disasters on cotton germination in the germination stage was predicted based on the model. In the seedling stage, the seedling yellowing rate was used as the response variable, and a linear regression model with the screened explanatory variables that significantly affected the seedling yellowing rate was constructed. The model was used to predict the impact of high temperature disasters on the development of cotton seedlings in the seedling stage. In the flowering stage, the floral deformation rate was taken as the response variable, and a linear regression model with the screened explanatory variables that significantly affected the floral deformation rate was constructed. The model was used to predict the impact of high temperature disasters on the development of cotton floral organs in the flowering stage of cotton. In the boll-setting stage, the cotton boll damage rate was taken as the response variable, and a linear regression model with significant explanatory variables that screened out the cotton boll damage rate was constructed. Based on this model, the impact of high temperature disasters on the maturity of cotton bolls in the boll-setting stage was predicted.

Images