Method for recommending economic fertilization of crops based on multi-source remote sensing data

Abstract

The invention discloses a crop economic fertilization recommendation method based on multi-source remote sensing data. The method comprises the following steps: establishing a nitrogen fertilizer effect curve by using a WOFOST model and an isoline; an economic nitrogen application amount equation is established, and the optimal economic nitrogen application amount is calculated; comprehensive crop growth monitoring indexes are made, and crop growth is graded; and distributing the nitrogen application amount according to a nutrient measurement and regulation method. According to the crop economic fertilization recommendation method based on the multi-source remote sensing data, aiming at the problem that corresponding yields under different nitrogen application amounts are difficult to obtain, a WOFOST model is provided, the nitrogen application amount is used as a unique variable, and the yield is recommended to be more accurate. The problem that corresponding yields are difficult to obtain under different nitrogen application amounts can be effectively solved, the advantages of a nitrogen fertilizer effect curve are obtained through a crop model, marginal cost is obtained according to the nitrogen fertilizer effect curve, a fertilization decision is made in combination with marginal income, farmland refined fertilization management can be achieved, the input cost is reduced, and the economic benefit is improved. And a new method is provided for refined farming operation of a large-area farmland.

Description

Technical Field

[0001] The present invention relates to the technical field of agriculture, and in particular to a method for recommending economic fertilization of crops based on multi-source remote sensing data. Background Art

[0002] By 2033, China's grain production will increase by 35% over the previous level to meet the growing food needs of the Chinese people. Nitrogen is a key factor affecting crop growth and development, and is also the main factor affecting its yield and aboveground biomass.

[0003] In recent years, due to blind fertilization and excessive nitrogen application in production, the utilization rate of nitrogen fertilizer has been low, which not only has failed to meet expectations, but has also caused environmental pollution and increased production costs. Summary of the invention

[0004] In view of the above-mentioned deficiencies in the prior art, the present invention provides a method for recommending economic fertilization of crops based on multi-source remote sensing data, which aims to solve the problems of blind fertilization and excessive nitrogen application in production, resulting in low nitrogen fertilizer utilization rate.

[0005] In order to achieve the above-mentioned invention object, the technical scheme adopted by the present invention is: a method for recommending economic fertilization of crops based on multi-source remote sensing data, including using the WOFOST model to establish a nitrogen fertilizer effect curve with equal lines; establishing an economic nitrogen application amount equation to calculate the optimal economic nitrogen application amount; making a comprehensive crop growth monitoring index to classify crop growth; and allocating the nitrogen application amount according to the nutrient measurement and control method;

[0006] Step 1: Using the WOFOST model, the nitrogen application rate was used as the only variable, and four fertilization levels (no fertilization, 0.5 conventional fertilization, conventional fertilization, and 1.5 conventional fertilization) were set. The nitrogen application rate ratio during the key growth period was combined to obtain the corresponding yield and establish the nitrogen fertilizer effect curve;

[0007] Step 2: Based on the nitrogen fertilizer effect curve in step 1, obtain the marginal cost, and establish the economic yield equation in combination with the marginal income (i.e., the unit price of crop sales). Substitute the obtained economic yield into the nitrogen fertilizer effect curve to obtain the optimal economic nitrogen application rate;

[0008] Step 3: Considering the contribution rate of agronomic parameters such as nitrogen concentration, chlorophyll concentration, plant height, and stem diameter combined with vegetation index, the comprehensive crop growth monitoring index is obtained, and the crop growth is divided into 7 levels based on statistical theories such as quartiles;

[0009] Step 4: Using the periodicity of remote sensing images, allocate the economic nitrogen application rate to each key growth period in proportion. Then, based on the 7 levels in step 3 and the nutrient measurement and control method, allocate the economic nitrogen application rate in the key growth period to each fertilization unit.

[0010] Furthermore, in step 1, based on the WOFOST model simulation, different nitrogen application rates are allocated to each key growth period to obtain corresponding crop yields, and step 1 further includes:

[0011] Step 1.1: According to the nitrogen application rate in the field survey, this nitrogen application rate is used as the conventional nitrogen application rate treatment, and four nitrogen fertilizer levels are set (no fertilization, 0.5× conventional nitrogen fertilizer, conventional nitrogen fertilizer, and 1.5× conventional nitrogen fertilizer), which are proportionally allocated to the basal fertilizer and each key growth period (such as basal fertilizer: greening period: jointing period: heading period = 0.3﹕0.2﹕0.3﹕0.2). The yield under different nitrogen application rates is obtained through the WOFOST crop model;

[0012] Step 1.2: Simulate the statistical model between yield and different nitrogen application rates to obtain a set of regression models between different nitrogen application rates and yield, and then use the modeling R 2 and predict R 2 The fitting effect of the evaluation model ranges from 0 to 1. The closer to 1, the higher the model accuracy. The most accurate one is the nitrogen fertilizer effect curve.

[0013] Furthermore, in step 2, the marginal cost is combined with the marginal income (i.e., the unit price of crop sales) to establish an economic nitrogen application rate equation, and step 2 further includes:

[0014] Step 2.1: Substitute nitrogen cost = nitrogen unit price × nitrogen application amount into the nitrogen fertilizer effect curve, calculate the inverse function of the nitrogen fertilizer effect curve, and then calculate the derivative of the inverse function, that is, the marginal cost;

[0015] Step 2.2: Microeconomic theory holds that the output when marginal cost equals marginal revenue is the output with the maximum profit for the enterprise. Based on this, the economic output equation is established;

[0016] Step 2.3: Substitute the economic yield obtained into the nitrogen fertilizer effect curve to obtain the optimal economic nitrogen application rate.

[0017] Furthermore, in step 3, considering that the importance of crop growth parameters in the comprehensive growth monitoring index is different, the dimensions of each index are not unified, etc., the weight of the growth index is determined by using the coefficient of variation method, and then the comprehensive growth monitoring index of the key growth period of the crop is constructed. Step 3 further includes:

[0018] Step 3.1: Download the Gaofen-1 image data with a spatial resolution of 2m corresponding to the key growth period, pre-process, and calculate the Normalized Difference Vegetation Index (NDVI), Ratio Vegetation Index (RVI), Enhanced Vegetation Index (EVI), Difference Vegetation Index (DVI), etc. The tools used mainly include ENVI, Python, etc.

[0019] Step 3.2: Use field measured data combined with the basic principle of coefficient of variation method to construct comprehensive growth monitoring indicators for crops during key growth periods;

[0020] Step 3.3: Using the regression random forest classification algorithm as the framework, perform parameter optimization, combine vegetation indices such as NDVI, RVI, EVI, and DVI, select the optimal parameters, and calculate the importance of the optimal parameters. Select 70% of the features as the training set and 30% as the test set. After the model is trained, predict the CGMI of the monitoring area;

[0021] Step 3.4: Obtain the number of pixels in the study area, arrange the comprehensive crop growth monitoring index CGMI in ascending order, and divide the crop growth into 7 levels based on statistical theories such as quartiles.

[0022] Furthermore, the 7 levels in step 3 are combined with the nutrient measurement and control method to allocate the economic nitrogen application amount to each fertilization unit, and step 4 further includes:

[0023] Step 4.1: Divide the fertilization units according to the operation parameters of the agricultural machinery, establish a fishing net with corresponding parameters, and assign the level to the fishing net through spatial connection;

[0024] Step 4.2: According to the spatial distribution of remote sensing growth, obtain the area proportion of different levels in the key growth period, and calculate the formula:

[0025]

[0026] Step 4.3: According to the remote sensing growth spatial distribution and area ratio, nitrogen ratio, obtain the nitrogen optimization index, calculation formula:

[0027] Nitrogen optimization index Yi = Si × i-grade nitrogen ratio

[0028] Step 4.4: According to the nitrogen horizontal optimization coefficient, obtain the horizontal nitrogen demand index, the calculation formula is:

[0029]

[0030] Step 4.5: Allocate the optimal economic nitrogen application rate to the base fertilizer and the key growth period of the crop in proportion (e.g. base fertilizer: greening period: jointing period: heading period = 0.3:0.2:0.3:0.2) to obtain the total nitrogen application rate for a certain key growth period;

[0031] Economic total nitrogen application rate N during the growth period 生育期 = Total economic nitrogen application amount × nitrogen fertilizer proportion in this growth period

[0032] Step 4.6, according to the horizontal nitrogen demand index, obtain the fertilizer amount at each level during the key growth period.

[0033] Fertilizer application amount at each level during the growth period = Ni ×N 生育期

[0034] The beneficial effects of the present invention are:

[0035] 1. The present invention provides a method for recommending economic fertilization of crops based on multi-source remote sensing data. Aiming at the problem that it is difficult to obtain the corresponding yield under different nitrogen application rates, a method for obtaining the corresponding yield is proposed based on the WOFOST model, with nitrogen application rate as the only variable, and four fertilization levels (no fertilization, 0.5 conventional fertilization, conventional fertilization and 1.5 conventional fertilization). The method can effectively solve the problem that it is difficult to obtain the corresponding yield under different nitrogen application rates.

[0036] 2. The method of recommending economic fertilization of crops based on multi-source remote sensing data of the present invention proposes an economic nitrogen application rate equation from an economic perspective, based on the economic theory that when the output increases to the point where the marginal cost is equal to the marginal income, the enterprise obtains its maximum profit. The economic nitrogen application rate equation is constructed; while pursuing higher output, input costs are taken into account to increase profits.

[0037] 3. The method of recommending economic fertilization for crops based on multi-source remote sensing data of the present invention takes into account the different importance of crop growth parameters in comprehensive growth monitoring indicators and the non-uniform dimensions of various indicators, and uses the coefficient of variation method to determine the weights of growth indicators, thereby constructing comprehensive growth monitoring indicators.

[0038] 4. The present invention provides a method for recommending economic fertilizer application for crops based on multi-source remote sensing data, which proposes a method based on growth grade division combined with nutrient measurement and control to distribute the economic nitrogen application amount to each fertilization unit, making the fertilization operation more refined, which is helpful for crop growth, increasing yield and reducing the waste of nitrogen fertilizer. BRIEF DESCRIPTION OF THE DRAWINGS

[0039] Figure 1 The present invention is a flow chart of a method for recommending economic fertilization of crops based on multi-source remote sensing data. DETAILED DESCRIPTION

[0040] In order to make the contents of the present invention more clearly understood, the technical solutions in the embodiments of the present invention will be clearly and completely described below in conjunction with the accompanying drawings in the embodiments of the present invention.

[0041] like Figure 1 As shown, a method for recommending economic fertilization of crops based on multi-source remote sensing data includes using the WOFOST model to establish a nitrogen fertilizer effect curve; establishing an economic nitrogen application rate equation to calculate the optimal economic nitrogen application rate; making a comprehensive crop growth monitoring index to classify crop growth; and allocating nitrogen application rates according to nutrient measurement and control methods;

[0042] Step 1: Using the WOFOST model, the nitrogen application rate was used as the only variable, and four fertilization levels (no fertilization, 0.5 conventional fertilization, conventional fertilization, and 1.5 conventional fertilization) were set. The nitrogen application rate ratio during the key growth period was combined to obtain the corresponding yield and establish the nitrogen fertilizer effect curve;

[0043] Step 2: Based on the nitrogen fertilizer effect curve in step 1, obtain the marginal cost, and establish the economic yield equation in combination with the marginal income (i.e., the unit price of crop sales). Substitute the obtained economic yield into the nitrogen fertilizer effect curve to obtain the optimal economic nitrogen application rate;

[0044] Step 3: Considering the contribution rate of agronomic parameters such as nitrogen concentration, chlorophyll concentration, plant height, and stem diameter combined with vegetation index, the comprehensive crop growth monitoring index is obtained, and the crop growth is divided into 7 levels based on statistical theories such as quartiles;

[0045] Step 4: Using the periodicity of remote sensing images, allocate the economic nitrogen application rate to each key growth period in proportion. Then, based on the 7 levels in step 3 and the nutrient measurement and control method, allocate the economic nitrogen application rate in the key growth period to each fertilization unit.

[0046] Through the above scheme, the specific experimental process is as follows:

[0047] (1) Download the Gaofen-1 satellite image with a spatial resolution of 2 m and perform preprocessing operations on the high-resolution image; perform radiometric calibration, atmospheric correction, registration, fusion, and cropping operations by using remote sensing data processing software such as ENVI.

[0048] (2) Obtain meteorological data within the monitoring area, mainly including: daily total radiation, daily maximum temperature, daily minimum temperature, rainfall, wind speed and other data during the entire growth period of crops in the previous year.

[0049] (3) Acquisition of soil parameters: Obtain soil parameter data in the monitoring area, mainly including soil texture, organic matter content, soil layer, soil moisture retention and other soil properties; match the soil data with the soil layer and depth required by the model, and ensure data quality and consistency.

[0050] (4) Obtain management practice data including irrigation, fertilization, sowing date, harvest date, etc. The data needs to match the time step of the model and be temporally coordinated with meteorological and soil data.

[0051] (5) Obtain agronomic parameters such as chlorophyll, nitrogen content, plant height, stem diameter, dry weight, and fresh weight of crops in the monitoring area.

[0052] (6) Obtain yields at different nitrogen application rates based on the WOFOST model. Configure the WOFOST model parameters based on the data in 2, 3, and 4, set the start and end time of the crop simulation, the crop type, and the configuration file, set the nitrogen application rate as the only variable, set four nitrogen fertilizer levels (no fertilization, 0.5× conventional nitrogen fertilizer, conventional nitrogen fertilizer, and 1.5× conventional nitrogen fertilizer), and distribute them to the base fertilizer and each key growth period in proportion. Obtain yields at different nitrogen application rates through the WOFOST crop model.

[0053] (7) Based on the statistical model between simulated yield and different nitrogen application rates in 6, a set of regression models between different nitrogen application rates and yield was obtained, and the fitting effect of the model was evaluated by modeling R2 and predicted R2, ranging from 0 to 1. The closer to 1, the higher the model accuracy.

[0054] Yield (y) = a × nitrogen application rate 2 +b×nitrogen application amount+c

[0055]

[0056] The coefficients a, b, and c in the formula are obtained by fitting the yield and different nitrogen application rates.

[0057] (8) Substituting nitrogen cost = nitrogen unit price × nitrogen application amount into the nitrogen fertilizer effect curve in formula 1, we get:

[0058]

[0059] (9) Find the inverse function of formula 3, and we get:

[0060]

[0061] (10) Taking the derivative of the nitrogen cost in equation (4), we get the marginal cost, which is as follows:

[0062]

[0063] (11) By setting marginal cost equal to marginal revenue (i.e., the selling price per unit of crop output), we can obtain the economic output equation, which is as follows:

[0064]

[0065] (12) According to the economic yield obtained in formula (6), substitute it into the nitrogen fertilizer effect curve in formula (1) to obtain the optimal economic nitrogen application rate.

[0066] (13) Calculate the Normalized Difference Vegetation Index (NDVI), Ratio Vegetation Index (RVI), Enhanced Vegetation Index (EVI), and Difference Vegetation Index (DVI) of remote sensing images in the key growth period monitoring area.

[0067]

[0068] (14) The weight is determined by the coefficient of variation method, and the calculation formula is as follows:

[0069]

[0070] Where V i is the coefficient of variation of the i-th indicator, σi is the standard deviation of the i-th indicator, is the average of the i-th indicator, W i is the weight of the ith indicator.

[0071] (15) Based on the 14 weighted data, the field measured data is combined with the basic principle of the coefficient of variation method to construct a comprehensive crop growth monitoring index. CGMI calculation formula:

[0072]

[0073] Where U i represents the normalized index of the i-th category, X i Represents the original i-th index, original index categories: chlorophyll, nitrogen content, plant height, stem diameter, dry weight, fresh weight, max(X i ) is the maximum value among the original i-type indicators.

[0074] (16) Using the regression random forest classification algorithm as the framework, parameter optimization was performed. The optimal parameters were selected by combining vegetation indices such as NDVI, RVI, EVI, and DVI. The importance of the optimal parameters was calculated. 70% of the features were selected as the training set and 30% as the test set. After the model was trained, the CGMI of the monitoring area was predicted.

[0075] (17) The optimal model of formula (15) is applied to the entire monitoring area. Combined with the crop planting range of the monitoring area, the crops in the entire monitoring area are inverted and mapped. At the same time, the roads in the monitoring area are removed. According to the model coefficients, Python is used for inversion to obtain the crop growth distribution map of the monitoring area. The better the growth of the crop area, the darker the color.

[0076] (18) Obtain the number of pixels in the study area, and arrange the CGMI in ascending order. The crop growth condition is divided into 7 levels based on statistical theories such as quartiles. The level division standard is as follows: CGMI level division standard for crop growth monitoring index

[0077]

[0078] (19) According to the operation parameters of agricultural machinery, the fertilization units are divided, and a fishing net with corresponding parameters is established. The levels are assigned to the fishing net through spatial connection.

[0079] (20) Calculate the area proportion of each level using the following formula:

[0080] Area ratio

[0081] (21) According to the area percentage data of formula 15, the nitrogen percentage is used to obtain the nitrogen optimization index. Calculation formula:

[0082] Nitrogen optimization index Yi = Si × i-grade nitrogen ratio

[0083] (22) According to the nitrogen optimization index Y of formula 16 i , and obtain the horizontal nitrogen demand index. Calculation formula:

[0084] Horizontal nitrogen demand index

[0085] (23) The optimal economic nitrogen application rate is proportionally distributed to the base fertilizer and the key growth period of the crop (e.g., base fertilizer: greening period: jointing period: heading period = 0.3:0.2:0.3:0.2) to obtain the total nitrogen application rate in a certain key growth period.

[0086] Economic total nitrogen application rate N during the growth period 生育期 = Total economic nitrogen application amount × nitrogen fertilizer proportion in this growth period

[0087] (24) Combined with the horizontal nitrogen demand index of formula 18, the fertilizer application amount at each level in a certain growth period is obtained. Calculation formula:

[0088] Fertilizer application amount at each level during the growth period = N i ×N 生育期

[0089] The present invention proposes a division based on growth grade combined with a nutrient measurement and control method to distribute the economic nitrogen amount to each fertilization unit, making the fertilization operation more refined, which is conducive to crop growth, increases yield, and reduces the waste of nitrogen fertilizer.

[0090] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions and improvements made within the spirit and principles of the present invention shall be included in the protection scope of the present invention.

Claims

1. A method for recommending economic fertilization of crops based on multi-source remote sensing data, characterized in that: It includes using the WOFOST model to establish the nitrogen fertilizer effect curve; establishing the economic nitrogen application rate equation to calculate the optimal economic nitrogen application rate; making comprehensive crop growth monitoring indicators to classify crop growth; and allocating nitrogen application rates based on nutrient measurement and control methods; Step 1: Using the WOFOST model, the nitrogen application rate was used as the only variable, and four fertilization levels (no fertilization, 0.5 conventional fertilization, conventional fertilization, and 1.5 conventional fertilization) were set. The nitrogen application rate ratio during the key growth period was combined to obtain the corresponding yield and establish the nitrogen fertilizer effect curve; Step 2: Based on the nitrogen fertilizer effect curve in step 1, obtain the marginal cost, and establish the economic yield equation in combination with the marginal income (i.e., the unit price of crop sales). Substitute the obtained economic yield into the nitrogen fertilizer effect curve to obtain the optimal economic nitrogen application rate; Step 3: Considering the contribution rate of agronomic parameters such as nitrogen concentration, chlorophyll concentration, plant height, and stem diameter combined with vegetation index, the comprehensive crop growth monitoring index is obtained, and the crop growth is divided into 7 levels based on statistical theories such as quartiles; Step 4: Using the periodicity of remote sensing images, allocate the economic nitrogen application rate to each key growth period in proportion. Then, based on the 7 levels in step 3 and the nutrient measurement and control method, allocate the economic nitrogen application rate in the key growth period to each fertilization unit.

2. The method for recommending crop economic fertilization based on multi-source remote sensing data according to claim 1, characterized in that: In step 1, based on the WOFOST model simulation, different nitrogen application rates are allocated to each key growth period to obtain corresponding crop yields. Step 1 further includes: Step 1.1: According to the nitrogen application rate in the field survey, this nitrogen application rate is used as the conventional nitrogen application rate treatment, and four nitrogen fertilizer levels are set (no fertilization, 0.5× conventional nitrogen fertilizer, conventional nitrogen fertilizer, and 1.5× conventional nitrogen fertilizer), which are proportionally allocated to the basal fertilizer and each key growth period (such as basal fertilizer: greening period: jointing period: heading period = 0.3﹕0.2﹕0.3﹕0.2). The yield under different nitrogen application rates is obtained through the WOFOST crop model; Step 1.2: Simulate the statistical model between yield and different nitrogen application rates to obtain a set of regression models between different nitrogen application rates and yield, and then use the modeling R 2 and predict R 2 The fitting effect of the evaluation model ranges from 0 to 1. The closer to 1, the higher the model accuracy. The most accurate one is the nitrogen fertilizer effect curve.

3. The method for recommending crop economic fertilization based on multi-source remote sensing data according to claim 2, characterized in that: In step 2, the marginal cost is combined with the marginal income (i.e., the unit price of crop sales) to establish the economic nitrogen application rate equation, and step 2 further includes: Step 2.1: Substitute nitrogen cost = nitrogen unit price × nitrogen application amount into the nitrogen fertilizer effect curve, calculate the inverse function of the nitrogen fertilizer effect curve, and then calculate the derivative of the inverse function, that is, the marginal cost; Step 2.2: Microeconomic theory holds that the output when marginal cost equals marginal revenue is the output with the maximum profit for the enterprise. Based on this, the economic output equation is established; Step 2.3: Substitute the economic yield obtained into the nitrogen fertilizer effect curve to obtain the optimal economic nitrogen application rate.

4. The method for recommending crop economic fertilization based on multi-source remote sensing data according to claim 3, characterized in that: In step 3, considering that the importance of crop growth parameters in the comprehensive growth monitoring index is different, the dimensions of each index are not uniform, etc., the weight of the growth index is determined by using the coefficient of variation method, and then the comprehensive growth monitoring index of the crop key growth period is constructed. Step 3 further includes: Step 3.1: Download the Gaofen-1 image data with a spatial resolution of 2m corresponding to the key growth period, pre-process, and calculate the Normalized Difference Vegetation Index (NDVI), Ratio Vegetation Index (RVI), Enhanced Vegetation Index (EVI), Difference Vegetation Index (DVI), etc. The tools used mainly include ENVI, Python, etc. Step 3.2: Use field measured data combined with the basic principle of coefficient of variation method to construct comprehensive growth monitoring indicators for crops during key growth periods; Step 3.3: Using the regression random forest classification algorithm as the framework, perform parameter optimization, combine vegetation indices such as NDVI, RVI, EVI, and DVI, select the optimal parameters, and calculate the importance of the optimal parameters. Select 70% of the features as the training set and 30% as the test set. After the model is trained, predict the CGMI of the monitoring area; Step 3.4: Obtain the number of pixels in the study area, arrange the comprehensive crop growth monitoring index CGMI in ascending order, and divide the crop growth into 7 levels based on statistical theories such as quartiles.

5. The method for recommending economic fertilization of crops based on multi-source remote sensing data according to claim 4, characterized in that: The seven levels of step 3 are combined with the nutrient measurement and control method to distribute the economic nitrogen application amount to each fertilization unit, and step 4 further includes: Step 4.1: Divide the fertilization units according to the operation parameters of the agricultural machinery, establish the fishing net with the corresponding parameters, and assign the level to the fishing net through spatial connection; Step 4.2: According to the spatial distribution of remote sensing growth, obtain the area proportion of different levels in the key growth period, and calculate the formula: Step 4.3: According to the remote sensing growth spatial distribution and area ratio, nitrogen ratio, obtain the nitrogen optimization index, calculation formula: Nitrogen optimization index Yi = Si × i-grade nitrogen ratio Step 4.4: According to the nitrogen horizontal optimization coefficient, obtain the horizontal nitrogen demand index, the calculation formula is: Step 4.5: Allocate the optimal economic nitrogen application rate to the base fertilizer and the key growth period of the crop in proportion (e.g. base fertilizer: greening period: jointing period: heading period = 0.3:0.2:0.3:0.2) to obtain the total nitrogen application rate for a certain key growth period; Economic total nitrogen application rate N during the growth period 生育期 = Total economic nitrogen application amount × nitrogen fertilizer proportion in this growth period Step 4.6, according to the horizontal nitrogen demand index, obtain the fertilizer amount of each level in the key growth period, the fertilizer amount of each level in the growth period = N i ×N 生育期 .

Images