Summer corn drought detection method and system

Abstract

The invention relates to the technical field of hydrology and meteorology, and discloses a summer corn drought detection method and system.The method comprises the steps that day-by-day precipitation and evaporation capacity data in a certain area are obtained, and the effective precipitation and the actual water demand of summer corn are obtained; obtaining a difference value between the effective precipitation amount and the actual water demand amount of the summer corn within a certain period of time; performing trend fitting on the difference value D < t > by using a smooth spline so as to replace a position parameter of Log-logistic distribution, and obtaining a time-varying position parameter; acquiring Log-logistic distribution of a time-varying time sequence according to the time-varying position parameters, estimating the parameters by using a probability weighted distance method PWMs based on empirical frequency, and standardizing drought detection indexes; the invention provides a new non-stationary drought index of the summer maize, which considers the requirements of different crops in growth stages on moisture and the difference of drought tolerance, and can realize day-by-day fine detection of the drought of the summer maize.

Description

Technical Field

[0001] This invention relates to the field of hydrometeorology, and in particular to a method for detecting drought in summer maize. Background Technology

[0002] The IPCC Sixth Assessment Report indicates a continued rise in global surface temperature, with a warming of 1.1°C compared to 1850-1990. Increased greenhouse gas emissions will lead to more extreme weather events. Under this scenario, further changes in the global water cycle will result in stronger precipitation, floods, and more severe droughts. Drought, as a frequent, long-lasting, and widespread meteorological disaster, has severely impacted global economic development, particularly agricultural production. Drought indices can quantitatively describe drought conditions and characterize the severity of drought occurrence, serving as the foundation and core of drought monitoring and assessment. In 2010, Vicente-Serrano established a new Standardized Precipitation Evapotranspiration Index (SPEI) based on the difference between precipitation and temperature evapotranspiration. SPEI not only considers the impact of temperature on drought severity but also retains the multi-scale and multi-spatial characteristics of the SPI. Studies by Sun Peng et al. have found that SPEI is relatively sensitive to temperature, overestimating drought intensity and duration when assessing future drought changes. NSPEI can compensate for the shortcomings of SPEI—its sensitivity to temperature and its bias in detecting non-stationary droughts—and more accurately reflect the drought situation. However, these commonly used drought indices are only applicable to assessing drought occurring in a single stage, and most studies predict and assess drought on monthly or longer timescales, making them insufficiently applicable to short-term timescales such as crop growing seasons and critical growth stages. Monthly monitoring shows that drought can range from no drought to severe drought within a month, and monthly or longer-term droughts clearly cannot accurately capture the timing of drought occurrence. Furthermore, many drought-related studies only consider factors such as precipitation and temperature, neglecting factors such as crop growth stages and ignoring the complexity of drought occurrence. Soil drought at any growth stage will lead to reduced maize yield, but drought from the tasseling to the milk stage has the greatest impact on yield, indicating significant differences in water requirements and drought tolerance at different developmental stages. Drought indices that do not consider crop growth stages and their demand patterns cannot accurately assess the impact of drought on crops, thus affecting the applicability of drought risk assessment. To accurately identify short-term droughts during crop growth and development and precisely capture crop drought conditions at each growth stage, it is urgent to monitor and issue early warnings based on day-scale drought indices developed from crop coefficients at each growth stage. Summary of the Invention

[0003] The purpose of this invention is to propose a method and system for detecting drought in summer maize, thereby solving the technical problem that the drought index estimated by existing technologies cannot accurately assess the impact of drought on crops, thus affecting the applicability of drought risk assessment.

[0004] The method includes the following steps: Step S1: Obtain daily precipitation and evaporation data for a certain region to obtain the effective precipitation and actual water requirement for summer corn; Step S2: Obtain the difference between the effective precipitation and the actual water requirement for summer maize within a certain period of time. ; Step S3, use smoothed splines to adjust the difference. D t A trend fit is performed to replace the location parameters of the Log-logistic distribution, thus obtaining the time-varying location parameters. ; Step S4, based on time-varying position parameters Acquiring time-varying The log-logistic distribution of the time series was used to estimate parameters using probability-weighted interval methods (PWMs) based on empirical frequencies, and drought detection indicators were standardized.

[0005] A drought detection system for summer maize includes: Data acquisition module: Acquires daily precipitation and evaporation data for a certain region to obtain the effective precipitation and actual water requirement for summer maize; Difference Acquisition Module: Obtains the difference between the effective precipitation and the actual water requirement for summer maize within a certain time period. ; Trend fitting module: uses smoothed splines to fit the differences D t A trend fit is performed to replace the location parameters of the Log-logistic distribution, thus obtaining the time-varying location parameters. ; Drought index detection module: based on time-varying location parameters Acquiring time-varying The log-logistic distribution of the time series was used to estimate parameters using probability-weighted interval methods (PWMs) based on empirical frequencies, and drought detection indicators were standardized.

[0006] The beneficial effects of this invention are: the method and system take into account the crop coefficients of summer maize at each growth stage, and use crop water requirement instead of potential evapotranspiration. Its daily-scale calculation can overcome the shortcomings of the coarse monthly SPEI time scale, and can more accurately detect drought conditions during the growth period of summer maize, thereby improving and compensating for the shortcomings of conventional drought index in detecting summer maize under non-stationary conditions. Attached Figure Description

[0007] Figure 1 This is a flowchart of the summer maize drought monitoring method; Figure 2It shows the spatial distribution of water demand during different growth stages in the Huang-Huai-Hai Plain. Figure 3 This shows the spatial distribution of water shortage during different growth stages in the Huang-Huai-Hai Plain. Figure 4 This is a verification and evaluation of the NSPEE-corn index. Detailed Implementation

[0008] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0009] Soil drought at any growth stage will lead to reduced maize yield, but drought from the tasseling stage to the milk stage has the greatest impact on yield, indicating that the water requirement and drought tolerance of crops vary greatly at different developmental stages. Drought indices that do not consider the crop growth stage and the demand patterns at each stage cannot accurately assess the impact of drought on crops, thus affecting the applicability of drought risk assessment. To accurately identify short-term droughts occurring during crop growth and development and precisely capture crop drought conditions at each growth stage, this application provides a drought detection method for summer maize. The method includes the following steps: Step S1: Obtain daily precipitation and evaporation data for a certain region to obtain the effective precipitation and actual water requirement for summer corn; Step S2: Obtain the difference between the effective precipitation and the actual water requirement for summer maize within a certain period of time. ; Step S3, use smoothed splines to adjust the difference. D t A trend fit is performed to replace the location parameters of the Log-logistic distribution, thus obtaining the time-varying location parameters. ; Step S4, based on time-varying position parameters Acquiring time-varying The log-logistic distribution of the time series was used to estimate parameters using probability-weighted interval methods (PWMs) based on empirical frequencies, and drought detection indicators were standardized.

[0010] Further, in step S1, daily precipitation and evaporation data for a certain region are obtained to determine the effective precipitation and actual water requirement for summer corn.

[0011] Step S1.1: First, the Penman-Monteith method is used to calculate the potential evapotranspiration (…). The calculation formula is as follows:

[0012] In the formula, Potential evapotranspiration, mm / d; ρ represents the slope of the saturated vapor pressure-temperature curve, in kPa / ℃; Net surface radiation, MJ / (m2·d), Soil heat flux; The daily average temperature (°C) The wind speed at a height of 2m is m / s; The saturated vapor pressure is kPa. The actual water vapor pressure is expressed in kPa. is the wet / dry constant, kPa / ℃.

[0013] Step S1.2, then calculate the actual water requirement of the crop.

[0014]

[0015] In the formula Water requirement for crops, mm / d; This represents the coefficient for summer maize crops.

[0016] Step S1.3: Finally, calculate the effective precipitation for summer maize. The formula for effective precipitation is as follows:

[0017] In the formula, Effective precipitation (mm / d) This represents the actual precipitation (mm / d).

[0018] The average water requirement of summer maize at each growth stage in the Huang-Huai-Hai Plain over the past 50 years has spatially shown a pattern of less water in the north and more in the south, generally varying along the latitudinal direction. Lower latitudes, higher temperatures, and greater evaporation rates result in higher water requirements. Figure 2 Sowing and seedling emergence period ( Figure 2 (a) represents the stage of maize growth with the lowest water requirement, with a multi-year average water requirement between 13.9 and 20.1 mm. Higher water requirements are observed in the southern part of the Huang-Huai-Hai Plain, primarily in northern Anhui Province and southeastern Henan Province. The multi-year average water requirement during the emergence and jointing stage is between 70.9 and 96.9 mm. Water requirements begin to increase at this stage, with a more significant increase in the central regions, indicating that maize requires more water during jointing. The areas with high water requirements remain in the southern part of the Huang-Huai-Hai Plain. Figure 2 (b) The water requirement during the jointing and tasseling stages and the tasseling-milk ripening stage exhibits a latitudinal zonation characteristic, with less water required in the north and more in the south. Looking at the entire growth stage of summer maize, the water requirement during the tasseling-milk ripening stage ranges from 92.3 to 138.1 mm. This stage requires a large amount of water for grain filling and is also the stage with the highest water demand. Figure 2 (d) The water requirements of summer maize at different growth stages exhibit a significant latitudinal zonation pattern. Figure 2 a-2f), water demand increases significantly starting from the jointing stage ( Figure 2 c).

[0019] Water shortage varies significantly across different growth stages. Figure 3 Sowing and seedling emergence period ( Figure 3 a) and the seedling emergence and jointing stage ( Figure 3 (b) The multi-year average water shortage shows a spatial characteristic of being high in the middle and decreasing gradually to the north and south, indicating that the effective precipitation in the central part of the study area cannot meet the water demand. The sowing and seedling emergence period is also the stage with the least water shortage in the entire growth process. (Jointing and tasseling stage...) Figure 3 (c) The multi-year average water deficit ranges from 30.4 to 78.2 mm, increasing spatially from north to south. During this stage, maize requires a large amount of water for jointing, and the water deficit is more severe in the southern part of the study area. The multi-year average water deficit for summer maize at the tasseling and milk-ripe stage ranges from 49.7 to 96.4 mm. Figure 3 (d) During this stage, the water shortage in summer maize is the most severe among the five growth stages. Milk-ripe maturity stage ( Figure 3 (e) The average annual water shortage ranges from 10.4 to 24.2 mm. Unlike other growth stages, the areas with higher water shortages during this stage are concentrated in the central and eastern parts of the Huang-Huai-Hai Plain. Overall, the water shortage during this stage is relatively small, only higher than during the sowing and seedling emergence period. Throughout the entire growth period, the spatial distribution of water shortage shows a trend of increasing from the north and south towards the center, with the most severe water shortage occurring in the central part of the Huang-Huai-Hai Plain. Figure 3 f in the middle.

[0020] Further, step S2 specifically includes: Calculate the difference between daily precipitation and water demand.

[0021]

[0022] In the formula, It is the difference between precipitation and actual evapotranspiration. The effective daily precipitation This represents the actual daily evaporation.

[0023] Further, step S3 specifically includes: Step S3.1: Use the smoothing splines function to determine the sequence data. Fitting linear or nonlinear trends, for Perform fitting.

[0024]

[0025] In the formula, Solar radiation; For time; For smoothing parameters; The highest temperature; The lowest temperature; for The linear fitting function.

[0026] Time-varying position parameters:

[0027] Furthermore, step S4 specifically includes: Based on time-varying The log-logistic distribution of the time series is as follows:

[0028] in, , , These are the scale, shape, and location parameters of the Log-logidic distribution function, which are estimated using probability-weighted interval PWMs based on empirical frequencies.

[0029] In the formula, It is an s-order PWM, where =4, It is the number of data points; The mean changes with the time series, and the trend value, i.e. the location parameter, fitted by the smooth spline function is constantly changing. Only when the mean remains unchanged will the location parameter remain unchanged. In this case, the NSPEI-corn and SPEI values ​​remain consistent. KS is used to determine whether it conforms to the Log-logistic distribution.

[0030]

[0031]

[0032] In the formula: For frequency estimation, when ≤0.5, for ;when >0.5, then Other parameters are: =2.515517, =0.802853, =0.01028, =1.432788, =0.189269, =0.001308, calculate NSPEI-corn. A positive NSPEI-corn value indicates wet conditions, and a negative NSPEI-corn value indicates dry conditions.

[0033] Figure 4 As shown, to evaluate the applicability of the constructed summer maize drought detection method, the daily NSPIE-corn and daily NSPIE index in the Huang-Huai-Hai Plain were input into seven machine learning models: Support Vector Machine (SVM), Lasso Regression, Random Forest, Ridge Regression, Stepwise Regression Analysis, and Decision Tree. The model was evaluated in conjunction with the actual yield reduction rate. The drought intensity and trend yield at different growth stages in the Huang-Huai-Hai Plain from 1970 to 2019 were used as input variables, and the yield reduction rate level was used as the output variable. All samples were randomly divided into training and test sets, with the training set accounting for 80% and the test set accounting for 20%. The prediction accuracy of all models—SVM, Lasso Regression, Random Forest, Ridge Regression, Stepwise Regression Analysis, and Decision Tree—was judged using the following indicators: Root Mean Square Error (RMSE), Mean Squared Error (MSE), Mean Absolute Percentage Error (MAPE), and Coefficient of Determination (R²). Figure 4 It can be seen that the daily NSPIE-Kc index performs better than the daily NSPIE index, indicating that the daily NSPIE-Kc index is closer to the actual drought process. The decision tree (DT) performs best among all evaluation indicators, with an R² value of 0.93, a mean absolute error (MAE) of 0.05, and a root mean square error (RMSE) of 0.19.

[0034] A drought detection system for summer maize includes: Data acquisition module: Acquires daily precipitation and evaporation data for a certain region to obtain the effective precipitation and actual water requirement for summer maize; Difference Acquisition Module: Obtains the difference between the effective precipitation and the actual water requirement for summer maize within a certain time period. ; Trend fitting module: uses smoothed splines to fit the differences D t A trend fit is performed to replace the location parameters of the Log-logistic distribution, thus obtaining the time-varying location parameters. ; Drought index detection module: based on time-varying location parameters Acquiring time-varying The log-logistic distribution of the time series was used to estimate parameters using probability-weighted interval methods (PWMs) based on empirical frequencies, and drought detection indicators were standardized.

[0035] The specific processing procedure of the data acquisition module is as follows: First, the potential evapotranspiration was calculated using the Penman-Monteith method. The calculation formula is as follows:

[0036] In the formula, Potential evapotranspiration, mm / d; ρ represents the slope of the saturated vapor pressure-temperature curve, in kPa / ℃; Net surface radiation, MJ / (m2·d), Soil heat flux; The daily average temperature is expressed in °C. The wind speed at a height of 2m is m / s; The saturated vapor pressure is kPa. The actual water vapor pressure is expressed in kPa. The constant of the wet and dry surface is kPa / ℃; Secondly, calculate the actual water requirements of the crop:

[0037] In the formula Water requirement for crops, mm / d; This refers to the summer maize crop coefficient; Finally, the effective precipitation for summer maize is calculated using the following formula:

[0038] In the formula, Effective precipitation is expressed in mm / d. This represents actual precipitation, expressed in mm / d.

[0039] The specific processing procedure of the difference acquisition module is as follows: Calculate the difference between daily precipitation and water demand.

[0040]

[0041] In the formula, It is the difference between precipitation and actual evapotranspiration. The effective daily precipitation This represents the actual daily evaporation.

[0042] The specific processing procedure of the trend fitting module is as follows: The sequence data was determined using the smoothing splines function. Fitting linear or nonlinear trends, for The fitting is performed as follows:

[0043] In the formula, Solar radiation; For time; For smoothing parameters; The highest temperature; The lowest temperature; for The linear fitting function; The time-varying position parameters are as follows: Loess (Locally Estimated Scatterplot Smoothing) is a nonparametric regression method. It combines local multinomial fitting and weighted regression to smooth data, and is particularly suitable for handling scatter plot data with nonlinear relationships.

[0044] The specific processing procedure of the drought index detection module is as follows: based on time-varying... The log-logistic distribution of the time series is as follows:

[0045] in, , , These are the scale, shape, and location parameters of the Log-logidic distribution function, which are estimated using probability-weighted interval PWMs based on empirical frequencies.

[0046] In the formula, It is an s-order PWM, where =4, It is the number of data points; As the mean changes with the time series, the trend value (i.e., the location parameter) fitted by the smoothed spline function is constantly changing. Only when the mean remains constant will the location parameter remain constant. Therefore, the NSPEI-corn and SPEI values ​​will be consistent. We can use KS to determine whether it conforms to a Log-logistic distribution.

[0047]

[0048] In the formula: For frequency estimation, when ≤0.5, for ;when >0.5, then Other parameters are: =2.515517, =0.802853, =0.01028, =1.432788, =0.189269, =0.001308, calculate NSPEE-corn. A positive NSPEE-corn value indicates a wet climate, while a negative NSPEE-corn value indicates a dry climate.

[0049] Finally, it should be noted that the above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art can still modify the technical solutions described in the foregoing embodiments or make equivalent substitutions for some of the technical features. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

Claims

1. A method for detecting drought in summer maize, characterized in that, The method includes the following steps: Step S1: Obtain daily precipitation and evaporation data for a certain region to obtain the effective precipitation and actual water requirement for summer corn; Step S2: Obtain the difference between the effective precipitation and the actual water requirement for summer maize within a certain period of time. ; Step S3, use smoothed splines to adjust the difference. D t A trend fit is performed to replace the location parameters of the Log-logistic distribution, thus obtaining the time-varying location parameters. ; Step S4, based on time-varying position parameters Acquiring time-varying The log-logistic distribution of the time series was used to estimate parameters using probability-weighted interval methods (PWMs) based on empirical frequencies, and drought detection indicators were standardized.

2. The method for detecting drought in summer maize according to claim 1, characterized in that, Step S1 specifically involves: Step S1.1: First, the potential evapotranspiration is calculated using the Penman-Monteith method. The calculation formula is as follows: In the formula, Potential evapotranspiration, mm / d; ρ represents the slope of the saturated vapor pressure-temperature curve, in kPa / ℃; Net surface radiation, MJ / (m2·d), Soil heat flux; The daily average temperature is expressed in °C. The wind speed at a height of 2m is m / s; The saturated vapor pressure is kPa. The actual water vapor pressure is expressed in kPa. The constant of the wet and dry surface is kPa / ℃; Step S1.2: Calculate the actual water requirement of the crop. In the formula Water requirement for crops, mm / d; This refers to the summer maize crop coefficient; Step S1.3: Finally, the effective precipitation for summer maize is calculated. The formula for effective precipitation is as follows: In the formula, Effective precipitation is expressed in mm / d. This represents actual precipitation, expressed in mm / d.

3. The method for detecting drought in summer maize according to claim 1, characterized in that, Step S2 specifically involves: Calculate the difference between daily precipitation and water demand. In the formula, It is the difference between precipitation and actual evapotranspiration. The effective daily precipitation This represents the actual daily evaporation.

4. The method for detecting drought in summer maize according to claim 1, characterized in that, Step S3 specifically involves: Step S3.1: Use the smoothing splines function to determine the sequence data. Fitting linear or nonlinear trends, for The fitting is performed as follows: In the formula, Solar radiation; For time; For smoothing parameters; The highest temperature; The lowest temperature; for The linear fitting function; The time-varying position parameters are as follows: 。 5. The method for detecting drought in summer maize according to claim 1, characterized in that, Step S4 specifically involves: Based on time-varying The log-logistic distribution of the time series is as follows: in, , , These are the scale, shape, and location parameters of the Log-logidic distribution function, which are estimated using probability-weighted interval PWMs based on empirical frequencies. In the formula, It is an s-order PWM, where =4, It is the number of data points; As the mean changes with the time series, the trend value (i.e., the location parameter) fitted by the smoothed spline function is constantly changing. Only when the mean remains constant will the location parameter remain constant. Therefore, the NSPEI-corn and SPEI values ​​will be consistent. We can use KS to determine whether it conforms to a Log-logistic distribution. In the formula: For frequency estimation, when ≤0.5, for ;when >0.5, then Other parameters are: =2.515517, =0.802853, =0.01028, =1.432788, =0.189269, =0.001308, calculate NSPEE-corn. A positive NSPEE-corn value indicates a wet climate, while a negative NSPEE-corn value indicates a dry climate.

6. A drought detection system for summer maize, characterized in that: include: Data acquisition module: Acquires daily precipitation and evaporation data for a certain region to obtain the effective precipitation and actual water requirement for summer maize; Difference Acquisition Module: Obtains the difference between the effective precipitation and the actual water requirement for summer maize within a certain time period. ; Trend fitting module: uses smoothed splines to fit the differences D t A trend fit is performed to replace the location parameters of the Log-logistic distribution, thus obtaining the time-varying location parameters. ; Drought index detection module: based on time-varying location parameters Acquiring time-varying The log-logistic distribution of the time series was used to estimate parameters using probability-weighted interval methods (PWMs) based on empirical frequencies, and drought detection indicators were standardized.

7. The summer maize drought detection system as described in claim 6, characterized in that: The specific processing procedure of the data acquisition module is as follows: First, the potential evapotranspiration was calculated using the Penman-Monteith method. The calculation formula is as follows: In the formula, Potential evapotranspiration, mm / d; ρ represents the slope of the saturated vapor pressure-temperature curve, in kPa / ℃; Net surface radiation, MJ / (m2·d), Soil heat flux; The daily average temperature is expressed in °C. The wind speed at a height of 2m is m / s; The saturated vapor pressure is kPa. The actual water vapor pressure is expressed in kPa. The constant of the wet and dry surface is kPa / ℃; Secondly, calculate the actual water requirements of the crop: In the formula Water requirement for crops, mm / d; This refers to the summer maize crop coefficient; Finally, the effective precipitation for summer maize is calculated using the following formula: In the formula, Effective precipitation is expressed in mm / d. This represents actual precipitation, expressed in mm / d.

8. The summer maize drought detection system as described in claim 6, characterized in that: The specific processing procedure of the difference acquisition module is as follows: Calculate the difference between daily precipitation and water demand. In the formula, It is the difference between precipitation and actual evapotranspiration. The effective daily precipitation This represents the actual daily evaporation.

9. A summer maize drought detection system as described in claim 6, characterized in that, The specific processing procedure of the trend fitting module is as follows: The sequence data was determined using the smoothing splines function. Fitting linear or nonlinear trends, for The fitting is performed as follows: In the formula, Solar radiation; For time; For smoothing parameters; The highest temperature; The lowest temperature; for The linear fitting function; The time-varying position parameters are as follows: Loess regression is a locally weighted regression method, which is a nonparametric regression method.

10. A summer maize drought detection system as described in claim 6, characterized in that: The specific processing procedure of the drought index detection module is as follows: based on time-varying... The log-logistic distribution of the time series is as follows: in, , , These are the scale, shape, and location parameters of the Log-logidic distribution function, which are estimated using probability-weighted interval PWMs based on empirical frequencies. In the formula, It is an s-order PWM, where =4, It is the number of data points; As the mean changes with the time series, the trend value (i.e., the location parameter) fitted by the smoothed spline function is constantly changing. Only when the mean remains constant will the location parameter remain constant. Therefore, the NSPEI-corn and SPEI values ​​will be consistent. We can use KS to determine whether it conforms to a Log-logistic distribution. In the formula: For frequency estimation, when ≤0.5, for ;when >0.5, then Other parameters are: =2.515517, =0.802853, =0.01028, =1.432788, =0.189269, =0.001308, calculate NSPEE-corn. A positive NSPEE-corn value indicates a wet climate, while a negative NSPEE-corn value indicates a dry climate.

Images