A precise management method for rice cultivation based on crop growth model

Abstract

The present invention provides a rice cultivation precision management method based on a crop growth model, and belongs to the technical field of agricultural management. After predicting the meteorological data of the season, the present invention sets different nitrogen application amounts and different planting densities according to the predicted meteorological data of the season, drives the crop growth model multiple times, obtains the crop yield under different nitrogen application amounts and different planting densities, calculates the marginal net benefit according to the formula, uses the marginal net benefit as a basis for judgment, determines the target yield of the season, and then obtains the optimal nitrogen application amount and planting density according to the target yield, and determines the influence of the planting density and nitrogen application amount on the rice yield and economic benefit, guides the planting plan by comparing the data such as yield, economic benefit and partial productivity of nitrogen fertilizer, and finally provides management strategy support for rice production.

Description

Technical Field

[0001] The present invention relates to the technical field of agricultural management, and in particular to a precise management method for rice cultivation based on a crop growth model. Background Art

[0002] The global population is expected to reach 9 billion by 2050. How to feed the growing population is a major challenge facing global agriculture. Precision agriculture has been rated as one of the most promising agricultural technologies in the 21st century and is one of the important means to ensure food security. In precision agriculture management, in addition to considering the spatial variation of soil nutrients, the difference in weather conditions between years should also be considered. In the context of meteorological changes, climate is more important than soil nutrients to rice yield and farmers' income. Studies have shown that by optimizing cultivation plans and adapting to climate change, rice yield can be effectively increased. Crop growth models and related decision support systems have been successfully applied by researchers in many countries around the world to crop cultivation management, yield estimation, climate change, suitability evaluation, soil carbon and nitrogen balance, benefit evaluation and other fields, including recommending the optimal nitrogen application rate suitable for local varieties. However, there are currently few studies that can clearly explain the relationship between different meteorological conditions, different densities, and different nitrogen application rates and yield and economic benefits of rice. Few studies have achieved higher yields by optimizing density and nitrogen application rates through cultivation methods that adapt to climate conditions. Summary of the invention

[0003] In view of this, in order to solve the above problems, the object of the present invention is to provide a precise management method for rice cultivation based on a crop growth model.

[0004] In order to achieve the above-mentioned object of the invention, the present invention provides the following technical solutions:

[0005] The precise management method of rice cultivation based on crop growth model has the following specific steps:

[0006] Acquire historical meteorological data, and predict the meteorological data of the current season based on the historical meteorological data combined with the meteorological data that has occurred in the current season;

[0007] According to the forecasted meteorological data of the season, different nitrogen application rates and different planting densities are set, and the crop growth model is driven multiple times to obtain different yields under different nitrogen application rates and different planting densities.

[0008] Construct a marginal net benefit formula for nitrogen application amount and seedling cost, use the marginal net benefit as the basis for judgment, determine the target yield for the season, and then obtain the optimal nitrogen application amount and planting density based on the target yield.

[0009] Preferably, the method for predicting the daily maximum temperature is: at intervals of 10 days, the maximum temperatures of each growth period of rice in different years are counted, and the upper quartile Q3, lower quartile Q1, mean and median of the maximum temperature are used as the predicted values ​​of the maximum temperature of each growth period of rice.

[0010] Preferably, the specific steps for predicting the seasonal meteorological data are as follows:

[0011] Determine the meteorological data for each year in history; divide the meteorological stages according to the three key growth nodes of rice, and divide each stage into normal year type, high temperature year type, and low temperature year type based on historical meteorological data. Divide the entire growth period into 27 meteorological types and determine the meteorological type to which each historical year belongs;

[0012] Obtain the actual measured data of the daily maximum temperature and daily minimum temperature that have occurred in the current year, compare them with the meteorological data of historical years, and use the meteorological data type of the closest historical year as the predicted meteorological type for the current year.

[0013] Preferably, the specific steps of the multiple driving crop growth model are:

[0014] Different nitrogen application rates were set with fixed step lengths. To improve the calculation speed, 10 different nitrogen application rates were set.

[0015] Set different sowing or transplanting densities with fixed step lengths. To simplify the calculation, set 10 different densities.

[0016] By combining different densities and different nitrogen application rates, 100 cultivation plans were generated, which were input into the crop growth model in turn. Then, combined with the predicted meteorological data of the season, 100 corresponding yield data were obtained.

[0017] Preferably, the marginal net benefit formula is:

[0018] MNR=Yield*Price—Nrate*Ncost—Seedling*Scost;

[0019] Among them, MNR is marginal net revenue, Yield is the simulated rice yield under the meteorological conditions predicted in the season, different nitrogen application rates, and different density conditions, in kg ha -1 , Price is the price of rice; Nrate is the amount of nitrogen applied, in kgha -1 , Ncost is the price of nitrogen fertilizer; Seeding is the transplanting or sowing density per hectare, and Scost is the cost of seeds or seedlings.

[0020] Preferably, the rice precision cultivation method based on the crop growth model further comprises:

[0021] The application amount of phosphorus fertilizer and potash fertilizer is determined as follows:

[0022] According to the results of MNR, the nitrogen application and sowing and transplanting schemes with the highest value can be selected, and the optimal target yield for the season can be determined;

[0023] After the autumn harvest of the previous season, the available phosphorus and available potassium contents in the soil are determined; the national standard method is preferably used for determining the available phosphorus and available potassium contents in the soil.

[0024] The target yield of the model data and the phosphorus and potassium nutrient content of the soil are used to obtain the optimal application amount of phosphorus and potassium fertilizers according to the target yield method; the target yield method is preferably calculated using nutrient management software.

[0025] Finally, before production, the target yield for the season, nitrogen application amount, potassium application amount, phosphorus application amount, transplanting or sowing density can be obtained, enabling detailed planning of pre-production.

[0026] Preferably, the three key growth nodes of rice are: effective tillering stage, ear growth stage and fruiting stage.

[0027] By adopting the above technical scheme, the present invention has the following beneficial effects: the present invention predicts the meteorological data of the season by combining the meteorological information such as temperature, radiation, rainfall, etc. with the meteorological data that has occurred in the season through historical meteorological data, and sets different nitrogen application amounts and different planting densities according to the predicted meteorological data of the season, drives the crop growth model multiple times, obtains different yields under different nitrogen application amounts and different planting densities, constructs the marginal net benefit formula of nitrogen application amount and seedling cost, determines the target yield of the season with the marginal net benefit as the judgment basis, and then obtains the optimal nitrogen application amount and planting density according to the target yield, and determines the influence of rice density and nitrogen application amount on rice yield and economic benefit, guides the planting plan by comparing data such as yield, economic benefit and nitrogen fertilizer partial productivity, and finally provides management strategy support for rice production. BRIEF DESCRIPTION OF THE DRAWINGS

[0028] Figure 1 This is a flow chart of the rice precision cultivation method based on the crop growth model of the present invention;

[0029] Figure 2 This is the distribution map of sunshine radiation, maximum temperature, minimum temperature and rainfall in Jiansanjiang area in the past 65 years;

[0030] Figure 3 is the relationship between the upper quartile, mean, median, and lower quartile maximum and minimum temperatures;

[0031] Figure 4 The average (blue), median (yellow), upper and lower quartiles (green, red) and extreme value distribution diagrams of the daily maximum and minimum temperatures;

[0032] Figure 5 The average (blue), median (yellow), upper and lower quartiles (green and red) and extreme value distribution of daily radiation;

[0033] Figure 6 It is a graph showing the relationship between temperature and radiation for different types of data;

[0034] Figure 7 The average (blue), median (yellow), upper and lower quartiles (green and red) and extreme value distribution of daily rainfall;

[0035] Figure 8 This is a flow chart of the precision management method of cold-region rice based on remote sensing and crop growth model;

[0036] Fig. 9 The yield of Longjing 31 cold-region rice simulated by CERES-Rice under different meteorological conditions, nitrogen application rates and density conditions;

[0037] Fig.10 The yield of Longjing 21 cold-region rice simulated by CERES-Rice under different meteorological conditions, different nitrogen application rates and different planting conditions;

[0038] Fig.11 The MNR simulated by CERES-Rice for Longjing 31 cold-region rice under different meteorological conditions, different nitrogen application rates and different density conditions;

[0039] Fig.12 The MNR simulated by CERES-Rice for Longjing 21 cold-region rice under different meteorological conditions, different nitrogen application rates and different density conditions;

[0040] Fig.13 is the N PFP simulated by CERES-Rice of Longjing 31 cold-region rice under different meteorological conditions, different nitrogen application rates and different density conditions;

[0041] Fig.14 is the N PFP simulated by CERES-Rice of Longjing 21 cold-region rice under different meteorological conditions, different nitrogen application rates and different density conditions;

[0042] Fig.15 Distribution diagram of yield (a), MNR (b) and N PFP (c) of cold-region rice Longjing 31 under PRM, FP and ROM treatments under meteorological conditions in different years from 1954 to 2019;

[0043] Fig.16Distribution diagram of yield (a), MNR (b) and N PFP (c) of cold-region rice Longjing 21 under PRM, FP and ROM treatments under meteorological conditions in different years from 1954 to 2019;

[0044] Fig.17 Partial nitrogen fertilizer productivity of rice for different farmers under various rice management systems. DETAILED DESCRIPTION

[0045] The present invention provides a method for precise rice cultivation based on a crop growth model, and the specific steps are as follows:

[0046] Acquire historical meteorological data, and predict the meteorological data of the current season based on the historical meteorological data combined with the meteorological data that has occurred in the current season;

[0047] According to the forecasted meteorological data of the season, different nitrogen application rates and different planting densities are set, and the crop growth model is driven multiple times to obtain different yields under different nitrogen application rates and different planting densities.

[0048] Construct a marginal net benefit formula for nitrogen application amount and seedling cost, use the marginal net benefit as the basis for judgment, determine the target yield for the season, and then obtain the optimal nitrogen application amount and planting density based on the target yield.

[0049] The crop growth model of the present invention is a computer software model that describes the quantitative relationship between environmental factors, cultivation management and crop growth, development and yield according to dynamic principles; the crop growth model is preferably a rice growth model, and more preferably a CERES-Rice model.

[0050] In the present invention, the historical meteorological data includes temperature, radiation and rainfall; the temperature includes the daily minimum temperature and the daily maximum temperature; the temperature, radiation and rainfall preferably include temperature, radiation and rainfall under normal meteorological conditions and extreme conditions. The present invention preferably simulates unknown meteorological data by replacing historical data. More preferably, the historical data is 65 years of historical data from 1954 to 2019. The daily maximum temperature of the present invention is preferably the upper quartile, lower quartile, mean or median of the temperature of the year in the historical years. The present invention preferably obtains the highest temperature of each meteorological type first, and then calculates the lowest temperature according to the highest temperature calculation equation.

[0051] In the present invention, under normal meteorological conditions, the average temperature of the historical years is preferably used as the daily maximum temperature forecast value of the year; under extreme meteorological conditions, the lower quartile Q1 of the temperature of the historical years is preferably used as the forecast value of the extreme low temperature weather of the year, and the upper quartile Q3 is used as the forecast value of the extreme high temperature weather of the year.

[0052] When forecasting radiation, in normal years, it is preferred to use the average temperature of the historical years as the daily maximum temperature forecast value of the year; under extreme meteorological conditions, the upper quartile (Q3) and lower quartile data (Q1) of the daily radiation value of the historical data are calculated to replace the high radiation and low radiation conditions under extreme radiation conditions, respectively. Q3 is the value at the top 25%, and Q1 is the value at 75%.

[0053] When forecasting rainfall, the average value is chosen as the prediction of rainfall for the season because rainfall is less regular than temperature and radiation.

[0054] The present invention determines the meteorological data of each historical year according to the above method, divides the meteorological stage according to the three key growth nodes of rice, divides each stage into normal year type, high temperature year type and low temperature year type according to the historical meteorological data, divides the entire growth period into 27 meteorological types, and determines the meteorological type to which each historical year belongs; in the present invention, the actual measured data of the daily maximum temperature and the daily minimum temperature that have occurred in the current year, or the weather forecast for 40 days are obtained, and compared with the meteorological data of the historical year, and the meteorological data type of the closest historical year is used as the predicted meteorological type of the current year. The three key growth nodes of rice are effective tillering stage, ear growth stage and fruiting stage, respectively. The information of the 40-day weather forecast is preferably the 40-day weather forecast provided by the National Meteorological Observatory. The effective tillering period of the present invention is preferably from May 20 to June 25, which is the vegetative growth stage of rice; the ear growth period of the present invention is preferably from June 25 to July 31, which is a part of the reproductive growth of rice; the fruiting period of the present invention is from July 31 to September 30.

[0055] After predicting the meteorological data of the season, the present invention sets different nitrogen application amounts and different planting densities according to the predicted meteorological data of the season, drives the crop growth model multiple times, and obtains different yields under the conditions of different nitrogen application amounts and different planting densities;

[0056] Preferably, 10 different nitrogen application rates are set, namely 20, 40, 60, 80, 100, 120, 140, 160, 180, and 200 kg ha -1 ;

[0057] Preferably, 10 different planting densities are set, namely 23, 24, 25, 26, 27, 28, 29, 30, 31, and 32 holes / m -2 .

[0058] The multiple-drive crop growth model is specifically as follows: different densities and different nitrogen application amounts are combined to generate 100 cultivation plans, which are input into the crop growth model in turn, and then combined with the predicted meteorological data of the season to obtain 100 corresponding yield data.

[0059] The present invention constructs a marginal net benefit formula of nitrogen application amount and seedling cost, uses the marginal net benefit as a judgment basis, determines the target yield of the season, and then obtains the optimal nitrogen application amount and planting density based on the target yield. The marginal net benefit formula is:

[0060] MNR=Yield*Price—Nrate*Ncost—Seedling*Scost;

[0061] Yield is the simulated rice yield under the predicted meteorological conditions of the season, different nitrogen application rates, and different density conditions, in kg ha -1 , Price is the price of rice; Nrate is the amount of nitrogen applied, in kg ha -1 , Ncost is the price of nitrogen fertilizer; Seeding is the transplanting or sowing density per hectare, and Scost is the cost of seeds or seedlings.

[0062] The present invention is based on the crop growth model rice precision cultivation method, and also includes the application amount of phosphorus fertilizer and potassium fertilizer determination, specifically:

[0063] According to the results of MNR, the nitrogen application and transplanting (or sowing) scheme with the highest value can be selected, and the optimal target yield for the season can be determined;

[0064] The available phosphorus and available potassium content of the soil were measured after the autumn harvest of the previous season according to the national standard method;

[0065] The target yield of the model data and the phosphorus and potassium nutrient content of the soil are used to obtain the optimal application amount of phosphorus and potassium fertilizers according to the target yield method; the target yield method is preferably calculated using nutrient management software.

[0066] Finally, before production, the target yield for the season, nitrogen application amount, potassium application amount, phosphorus application amount, and transplanting (or sowing) density can be obtained, enabling detailed planning of pre-production.

[0067] Example 1 Determination of daily maximum and minimum temperature forecast data in historical years

[0068] Daily maximum and minimum temperatures are key parameters driving rice development in the CERES-Rice model. An analysis of meteorological data from 1954 to 2019 found that daily maximum and minimum temperatures showed strong interannual consistency over time, with the overall pattern of increasing first, reaching the highest value, and then decreasing. Therefore, it is feasible to use the quartile method and the mean and median values ​​to implement the forecast simulation scheme for meteorological data.

[0069] The experiment of the embodiment of the present invention was carried out in the Qixing Farm in Jiansanjiang, Heilongjiang Province. The actual values ​​and predicted values ​​of the highest and lowest temperatures at intervals of 10 days in each growth period were statistically analyzed. The results show that the effective accumulated temperature is closer to the actual value, and it is more reasonable to use it as the forecast data of the highest and lowest temperatures. However, the data input into CERES-Rice are the highest and lowest temperatures. If the conventional processing method is used, the highest and lowest temperatures need to be combined in pairs to obtain a temperature sequence. However, in the CERES-Rice model, the default lowest temperature must be lower than the highest temperature. Therefore, when setting the lowest temperature, the value of the highest temperature must be considered. In addition, considering that the lowest and highest temperatures are at the three key nodes of the rice growth period, both have normal values ​​and extreme values, if the conventional processing is used, the temperature needs to be set for 27 treatments, and the scenario simulation is extremely complex. In order to facilitate the processing of the forecast data of the highest and lowest temperatures, the present invention first explores the relationship between the upper quartile, lower quartile, mean, and median highest and lowest temperatures, and tries to use the highest temperature to calculate the lowest temperature. The results are as follows. Figure 3 shown.

[0070] Figure 3 It shows that no matter the upper quartile, the lower quartile, the mean, or the median, after exploration, it is found that there is a very significant correlation between the daily maximum temperature and the daily minimum temperature, and the R 2 Close to 1, the RMSE of the four types of data is also close to 1°C, so the highest temperature can be used to calculate the lowest temperature and replace the real lowest temperature. In the specific scenario simulation, the highest temperature of each type is first obtained, and then the lowest temperature is calculated according to the calculation equation based on the highest temperature. The specific formula is that the upper quartile is = 0.9616x-10.405, the average formula is y = 0.987x-10.264, the median lowest temperature calculation formula is y = 0.9863x-10.142, and the lower quartile lowest temperature calculation formula is y = 1.0068x-9.9131. Since the highest temperature and the lowest temperature are highly correlated, it is feasible to convert the highest temperature into the lowest temperature through the formula.

[0071] The quartile method was used to calculate the temperature data for 65 years. Figure 4 As shown, the larger part of the 65-year historical statistical data, i.e., the upper quartile (Q3), and the smaller part, i.e., the lower quartile value (Q1), are used as the extreme high temperature and extreme low temperature conditions, respectively. In order to explore which data, the median or the average, is more in line with the forecast requirements, this embodiment counted the actual maximum temperature data and the predicted maximum temperature data at intervals of 10 days during the entire growth period of rice from 2014 to 2018, and the statistical results are shown in Table 1.

[0072] Table 1 Relationship between the actual and predicted maximum temperature during the rice growth period in Jiansanjiang area (℃)

[0073]

[0074] Table 1 shows that compared with the highest temperature obtained by the median, the forecast value replaced by the average value is closer to the true value. Selecting the average value as the forecast value of unknown meteorological data can better reflect the actual temperature situation. Therefore, in the scenario simulation, the average value or median of the highest temperature in 65 years is used instead in normal years. Since extreme weather occurs in some years, this embodiment uses the result of quartile operation, and the upper quartile Q3 is used as the forecast value of extreme high temperature weather, and the lower quartile Q1 is used as the forecast value of extreme low temperature weather. By statistically analyzing the highest temperature of meteorological data in the test years 2014-2018, it is found that the average value of the highest temperature in these 5 years is higher than the forecast value of Q1, and the highest temperature in these 5 years is lower than the forecast value of Q3. Therefore, it is reasonable to use Q1 and Q3 as weather forecasts for extreme low and high temperatures. In addition, the daily minimum temperature forecasted in the years of extreme low and high temperatures is calculated using the above formula.

[0075] Example 2 Determination of solar radiation forecast data

[0076] Solar radiation is a necessary parameter for crop photosynthesis and an important input parameter in the CERES-Rice model. The present invention analyzes 65 years of radiation data and finds that the solar radiation data is consistent with the temperature data, and shows a trend of first increasing and then decreasing as the day progresses. The method for determining the solar radiation value in a normal year is consistent with the temperature determination method. The actual value of the solar radiation data during the critical period of the whole growth period of rice from 2014 to 2018 is compared with the average value and median value, as shown in Table 2.

[0077] Table 2 Relationship between solar radiation value and forecast value during rice growth period in Jiansanjiang area (MJ / m 2 )

[0078]

[0079]

[0080] The above results show that, similar to the maximum temperature, the average sunshine radiation value of Jiansanjiang Qixing Farm in 65 years is slightly different from the sunshine radiation value in the experimental years from 2014 to 2018. Therefore, it is more reasonable to select the average sunshine radiation value of 65 years as the forecast value of sunshine radiation. Similarly, for extremely high and extremely low sunshine radiation values, the upper quartile (Q3) and lower quartile (Q1) of the daily radiation values ​​of historical data are calculated to replace the high sunshine radiation and low sunshine radiation conditions under extreme sunshine radiation conditions. In the end, the sunshine radiation values ​​at the key nodes of the entire growth period of rice from 2014 to 2018 were higher than Q1 and lower than Q3. Therefore, the sunshine radiation value setting method obtained by this method is more reasonable. The final data is as follows Figure 5 shown.

[0081] After exploration, it was found that due to the high consistency of temperature and solar radiation in the daily time series over the past 65 years, high temperatures are often accompanied by strong solar radiation, while low temperatures are often accompanied by little solar radiation. Solar radiation and temperature are highly correlated. If the radiation can be calculated using the highest temperature, it will effectively reduce the generation of incorrect meteorological data. After trial and error, this embodiment splits the temperature into the first 200 days and the last 200 days, and the effect of establishing the correlation separately is the best, such as Figure 6 As shown. For the upper quartile, mean and median values, the highest temperature radiation has a good correlation, and the R 2 All of them are greater than 0.95, except for the median RMSE of 1-200 days, which is 1.1 MJ / m 2 , the RMSE of other types of data is less than 1MJ / m 2 Therefore, the maximum temperature can be used to obtain the solar radiation data through the quadratic equation. As for the relationship between the maximum temperature of the lower quartile and the solar radiation, there is a correlation between the two, R 2 The values ​​are 0.71 and 0.81 respectively, and the RMSE are 1.58 and 1.22 MJ / m 2 Although the effect is worse than the correlation of other types of data simulation, it is still within an acceptable error range. Therefore, the present invention simulates the sunshine radiation data with the highest temperature, and it is feasible to carry out meteorological research. In addition, the current mainstream meteorological forecast products in China only mention the forecast of the highest temperature, the lowest temperature and rainfall, and there are few sunshine radiation forecast products. This method can provide a basis for subsequent work.

[0082] Example 3: Determination of daily rainfall forecast data

[0083] Compared with temperature and solar radiation values, rainfall has poor regularity and is currently difficult to predict in the medium and long term. It is extremely complex, such as Figure 7As shown. In order to explore the impact of rainfall on yield and the response mechanism of smart solutions, this embodiment, based on the research of predecessors, uses the average daily rainfall of historical data as forecast data in normal years, and uses 1.5 times the average rainfall to replace the extreme rainfall in Q3 in extremely rainy years, and uses 0.5 times the daily average rainfall to replace the extreme rainfall in Q1 extreme rainy years to obtain rainfall data in extremely rainy years. The final result shows a special situation where there is rainfall every day, which is inconsistent with the actual production situation. In addition, since rice is in a flooded state for a long time and is not sensitive to irrigation water supplemented by rainwater, the CERES-Rice model is also insensitive to the impact of rainfall on rice yield. Therefore, this embodiment does not discuss the impact of rainfall on yield separately. The rainfall data obtained in this embodiment is mainly used to carry out research on precise management of rice under the conditions of weather forecast. Although the forecast results are quite different from the actual production situation, for the normal operation of the crop growth model, the present invention selects the mean as the forecast of rainfall in the season.

[0084] Table 3 Relationship between actual and forecast rainfall during the growth period of rice in cold regions (mm)

[0085]

[0086] Example 4 Economic Benefit Evaluation Method

[0087] The CERES-Rice model can output yield data. This embodiment will set different nitrogen application rates and transplanting densities to obtain yields under different meteorological conditions. At present, the main economic indicator for evaluating different management measures, Marginal Net Return (MNR, yuan), is a common benefit evaluation method, which refers to the total yield minus the cost. In order to evaluate the effects of nitrogen application rate and planting density on yield, economic benefits and nitrogen fertilizer partial productivity of cold-region rice under 27 different simulated meteorological conditions, a marginal net return calculation formula that comprehensively considers nitrogen application rate and seedlings is set:

[0088] MNR=Yield*Price-Nrate*Ncost-Seedling*Scost (Formula 1)

[0089] Yield is the simulated rice yield under different meteorological conditions, different nitrogen application rates, and different density conditions, and the unit is kg·ha -1 , Price is the price of rice. For the convenience of statistics and comparison, this paper assumes that the price of rice is constant at 2.96 yuan·kg -1 ,; Nrate is the nitrogen application rate, unit is kg·ha -1Ncost is the price of nitrogen fertilizer. In order to facilitate calculation, statistics and comparison, the present invention sets the price of nitrogen fertilizer to a fixed value of 3.5 yuan kg -1 ; Seeding is the transplanting density per hectare. In the present invention, the number of rice plants per hole for cold-region rice is uniformly set to 6 according to planting experience. According to market research, the selling price of each seedling is about 0.01 yuan. The final result will show the marginal net return under different conditions. Similarly, the optimal nitrogen application amount and density management strategy can be determined under different temperature and radiation conditions according to the maximum value of MNR.

[0090] Example 5 Precision Management Method Setting

[0091] In order to achieve nutrient and cultivation management in the whole process before and during production, the present invention proposes a set of precise management methods based on remote sensing and crop growth model coupling, which can obtain the optimal nitrogen, phosphorus, potassium fertilizer and density scheme before production, and the optimal ear fertilizer recommendation method during production. The workflow is as follows: Figure 8 As shown. The main process is to use historical meteorological data before production, according to the above-mentioned daily maximum temperature, daily minimum temperature, solar radiation, and daily rainfall forecast methods, to determine the annual meteorological data. Then, according to the 40-day daily maximum temperature and daily minimum temperature forecast data provided by the National Meteorological Observatory, and then according to the radiation forecast algorithm, determine the 40-day daily radiation data and update the meteorological data for the season. By setting 20-200kg·ha -1 , based on 10kg·ha -1 The CERES-Rice model was calibrated after years of field trials to simulate the achievable yield potential, the optimal planting density and nitrogen fertilizer application rate. The target yield of the model data and the phosphorus and potassium content of the soil were used, and the nutrient management software developed by the inventor was used to obtain the optimal phosphorus and potassium fertilizer application rates, of which 60% of the nitrogen fertilizer was applied using the side deep fertilization scheme, and the remaining 40% was used as ear fertilizer.

[0092] Experimental Example 1 Effects of nitrogen application rate and density on rice yield in cold regions under different temperature conditions

[0093] In order to explore the effects of different nitrogen application rates and different densities on yield under different meteorological conditions, obtain the optimal yield in different years, and provide a basis for precise management strategies. Since meteorological data are daily data, the data volume is large and the full analysis is difficult, the present invention is based on the growth and development law of cold-region rice in Jiansanjiang, and simplifies the difficulty of the research. Three different time periods are set, namely the early stage (May 20-June 25), which is the effective tillering stage and also the nutritional growth stage of rice; the middle stage (June 25-July 31), which is the ear growth stage and is part of the reproductive growth of rice; and the late stage (July 31-September 30), which is the fruiting stage. The three time periods are matched and combined with normal temperature (M), as well as Q1 and Q3, to obtain a total of 27 groups (3 3 ) Different meteorological data types. By continuously iterating and running different types of meteorological data, setting different densities and different nitrogen application rates, and running CERES-Rice, we finally obtained the yield data of two different varieties, Longjing 31 and Longjing 21, as shown below: Fig. 9 and Fig.10 shown.

[0094] Fig. 9 and Fig.10 The results show that for different nitrogen application rates, the yield of both varieties showed a trend of increasing with nitrogen application rate. The yield increased rapidly at the beginning, but the growth rate of yield slowed down with the application of nitrogen fertilizer. When the optimal nitrogen application rate was reached, the yield no longer increased. However, the optimal nitrogen application rate was different for different temperature years. Under the conventional planting density of 28 holes per square meter, from low temperature to high temperature, the optimal nitrogen application rate of Longjing 31 ranged from 90kg ha -1 ~150kg·ha -1 The yield of cold-region rice varies greatly under different meteorological conditions, corresponding to the amount of nitrogen applied. The yield range of Longjing 31 is 3.6 kg ha -1 ~11kg·ha -1 For Longjing 21, the optimal nitrogen application rate is 70 kg ha -1 ~140kg·ha -1 , the yield range is 3.3kg·ha -1 ~12kg·ha -1 .

[0095] For different densities, under early low temperature and early normal temperature conditions, rice yield showed a trend of increasing with the increase of cultivation density. However, under early high temperature conditions, the increase in density had little effect on yield. Under nitrogen deficiency conditions, no matter how the density was increased, the change in yield was not large. Under nitrogen sufficient conditions, early low temperature or normal temperature can effectively increase yield, but under early high temperature conditions, no matter how the nitrogen application rate changed, the increase in yield was not obvious.

[0096] Experimental Example 2 Effects of nitrogen application rate and density on MNR of rice in cold regions under different temperature conditions

[0097] In order to evaluate the effects of different nitrogen application rates and different densities on cold-region rice under different temperature conditions, the present invention introduces MNR to explore the effects of the optimal nitrogen application rate and density of Longjing 31 and Longjing 21 on the economic benefits of cold-region rice under different meteorological conditions. The results are as follows Fig.11 and Fig.12 shown. Fig.11 and Fig.12 The results show that MNR and rice yield show the same pattern with different density and nitrogen application rate. With the continuous increase of nitrogen fertilizer and seedling input, MNR increases accordingly. However, with the continuous increase of input, the growth rate of MNR gradually slows down. After reaching the optimal value, MNR will not change under most conditions, and MNR will decrease in some cases. Due to the low cost of nitrogen fertilizer and seedling input per hectare, the decline of MNR is not large.

[0098] For different nitrogen application rates, under different meteorological conditions and density conditions, MNR increases with the increase of nitrogen application. When the MNR reaches the optimal level, the MNR remains unchanged or decreases slightly. The optimal MNR is from 17 kg ha -1 ~31kg·ha -1 When in the early low temperature state, the overall MNR is not high due to the reduction in yield. The MNR can be effectively improved by appropriately increasing the density and nitrogen application, but it will no longer increase when it reaches the critical value. In the early low temperature period, it is easier to reach the critical value than under other meteorological conditions, indicating that under this meteorological condition, the overall yield is poor. Farmers should control inputs and obtain the maximum MNR under the optimal density and nitrogen application. Under the condition of continuous low temperature in the early, middle and late stages, the lowest MNR is 17 kg ha for Longjing 31. -1 The MNR of Longjing 21 was even lower, at 10 kg ha -1 , indicating that the 11-leaf variety Longjing 31 is more cold-resistant than the 12-leaf variety Longjing 21. Low temperatures in the early stage but normal or high temperatures in the middle and late stages can effectively improve the MNR of cold-region rice, with the highest MNR reaching 26 and 16 kg ha -1 Under normal climate conditions in the mid-term, the MNR can be increased rapidly by increasing the nitrogen application rate and density. At this time, under the conditions of continuous low temperature in the mid-to-late period and continuous high temperature in the mid-to-late period, the MNR of rice is as follows: Longjing 31 is 22-29 kg ha -1 Helongjing 21 is 22-29kg·ha -1Under this condition, if you want to achieve a larger MNR, you need to invest more nitrogen fertilizer and density. For early high temperature, MNR can reach the highest value, among which Longjing 31 reaches 32,000 yuan, and Longjing 21 can reach 31,000 yuan.

[0099] For different densities, increasing density can effectively increase MNR when nitrogen application is low, but when density increases to the maximum, MNR decreases. This is because there are too many rice plants in the early stage, which consumes a lot of nutrients in the soil, and the lack of fertilizer is serious in the later stage, resulting in a decrease in yield, and the increase in cost leads to a decrease in MNR. When nitrogen fertilizer is sufficient, MNR gradually increases with the increase in density, but the rate of increase slows down. Under low temperature conditions, the higher the density, the higher the corresponding MNR, while under high temperature conditions, MNR is generally best at 28 holes per square meter. In short, under different climatic conditions, with the increase in density and nitrogen application, the MNR of cold-region rice continues to increase, but the growth rate slows down until it finally does not change. To carry out precise management, it is necessary to optimize the most appropriate density and nitrogen application amount for different meteorological conditions to achieve the best benefits.

[0100] Experimental Example 3 Effects of nitrogen application rate and density on NPFP of rice in cold regions under different temperature conditions

[0101] While taking into account the high yield and high economic benefits of cold-region rice, it is necessary to consider resource utilization in order to reduce the environmental cost. Nitrogen fertilizer partial productivity (NPFP) is an indicator that describes the efficiency of nitrogen fertilizer utilization. It refers to the crop grain yield that can be produced by nitrogen fertilizer applied per unit area, that is, yield / nitrogen application amount. It is an important indicator that reflects the comprehensive effect of soil basic fertility and crop on fertilizer application in the test area. In order to explore the effects of nitrogen application amount and density on the NPFP of Jiansanjiang cold-region rice under different temperature conditions, the present invention set two different varieties in 27 different meteorological conditions, 10 densities and 10 nitrogen application amounts, and ran the CERES-Rice model to output the NPFP, such as Fig.13 and Fig.14 shown.

[0102] Since no nitrogen treatment was applied, NPFP could not be calculated. In order to facilitate comparative analysis, the NPFP in this paper was set to 500 kg kg -1. Under different meteorological conditions, the two varieties showed a trend of decreasing NPFP with increasing nitrogen application. With the increase of density, under low temperature conditions in the early stage, if the nitrogen application rate is too low, NPFP decreases with increasing density, and if the nitrogen application rate is high, NPFP increases with increasing density. Under normal temperature conditions in the early stage, NPFP is not affected by density under low nitrogen levels, but under high nitrogen conditions, NPFP increases with increasing density. Under high temperature conditions in the early stage, NPFP basically does not change with changes in density. For the Longjing Rice 31 variety, the maximum value of NPFP under the simulated meteorological conditions of high temperature in the early stage, low temperature in the middle stage, and normal in the late stage was 450kg·kg -1 The minimum NPFP value was 19 kg kg under the conditions of low temperature in the early stage, high temperature in the middle stage, and low temperature in the late stage. -1 For Longjing 21 rice, 440 kg kg -1 The N PFP of Longjing 31 is consistent with that of Longjing 31. Under the conditions of low temperature in the early stage, high temperature in the middle stage, and low temperature in the late stage, 17 kg kg -1 NPFP. Since NPFP is affected by yield and nitrogen application, it has nothing to do with density. Therefore, to obtain higher NPFP, the nitrogen application must be lower, and the yield must be high. In general, no matter what the conditions are, the lowest NPFP is in the treatment with the highest nitrogen application. Therefore, in order to achieve the best economic benefits and the lowest environmental costs, first select the combination with the highest MNR, and then screen out the treatment with the highest NPFP density and nitrogen application, so that you can obtain the most economic and environmentally optimal precise management plan under different meteorological conditions.

[0103] Experimental Example 4: Evaluation of the effect of antenatal planning based on precise management

[0104] In order to compare the effects of farmers' conventional treatment (FP), regional optimization treatment (ROM) and precision management prenatal planning (PRM), especially under different meteorological conditions caused by climate change, which management can obtain higher yields and higher benefits, and effectively improve the efficiency of nitrogen fertilizer utilization, the present invention uses meteorological data from 1954 to 2019 as the basis to carry out inter-annual statistics. In order to facilitate the comparative work of research and select the optimal management. The present invention first selects MNR as the basic comparison object. The highest MNR value under different densities and different nitrogen application rates is used as the first judgment basis. If the MNR values ​​are the same, NPFP is compared. Since MNR is simulated by the model, the model has an error of about 10%. Therefore, the present invention divides MNR by 1000 to eliminate the error, and takes the value downward to obtain a 2-digit MNR. The final result is as follows. Fig.15 and Fig.16As shown, it can be observed from the time series chart of 65 years that for Longjing 21 and Longjing 31, in terms of yield, PRM>FP>ROM, in terms of MNR, PRM>FP>ROM, and for NPFP, ROM>PRM>FP. It can be seen that PRM management performs well in terms of yield and MNR, but is slightly worse than ROM in terms of NPFP.

[0105] For Longjing 31, the optimal nitrogen application rate range is 80-160 kg·ha -1 The optimal density is between 24 and 32 holes per square meter, and the final yield in 65 years is 561 tons, with an MNR of 1.265 million yuan and an average NPFP of 61 kg kg -1 , distributed in 32-114kg kg -1 For farmers’ conventional treatment (FP), the yield in 65 years was 526 tons, the MNR was 1.205 million yuan, and the average yield of NPFP in 65 years was 53kg kg -1 For ROM management, 65 years of continuous production can generate a production of 524 tons and obtain MNR of 1.201 million yuan, but the N PFP is 64.4 kg kg -1 In general, in Longjing 31, precision management of prenatal planning (PRM management) can increase yield by about 6.7% and increase MNR by about 60,000 yuan, an increase of 5%, compared with farmers' conventional treatment (FP). At the same time, N PFP is 8kg kg higher than farmers' conventional treatment (FP). -1 , an increase of about 15%. Compared with ROM management, PRM can increase production by 7.1% and MNR by 5.3% over the 65-year time series, but for NPFP, PRM is 3.4 kg kg lower than ROM -1 .

[0106] The PRM management of Longjing 21 is similar to that of Longjing 31, and the optimal nitrogen application rate is 61-160 kg·ha -1 The optimal density is between 23 and 32 holes per square meter. The optimal yield can reach 537.3 tons in 65 years, with an MNR of 1.197 million yuan. The average NPFP in 65 years is 57.7 kg kg -1 Compared with PRM and FP, 502.9 tons of rice can be produced in 65 years, which is 6.8% more, and the MNR is 1.148 million yuan, which is 4.5% more, and the NPFP of farmers is 50.7 kg kg -1 , increased by 7kg kg -1 , an increase of 13.8%. Comparing PRM and ROM, ROM can produce about 498 tons of rice in 65 years, an increase of about 7.9%, MNR increased by 77,000 yuan, an increase of 6.9%, and ROM's NPFP is 60.3kg·kg-1 , PRM is 2.6kg·kg lower than ROM -1 .

[0107] In summary, the results of multi-year simulations of the effects of different varieties of cold-region rice were conducted from 1954 to 2019. The results showed that Longjing 21 and Longjing 31 could effectively increase yield and MNR by using the PRM method compared with farmers and ROM management. In addition, the NPFP was 8-15% higher than that of farmers, but 2-3 kg·kg lower than that of ROM. -1 PRM can adjust according to the changes in the weather over a long period of time and select the appropriate density and nitrogen application rate, which can achieve the highest yield and the best benefits compared with ROM and FP.

[0108] Experimental Example 5 Field Test

[0109] The field experiment was conducted in Jiansanjiang Qixing Farm in 2018. Through farmer surveys, the yields in the past two years were investigated. Finally, four farmers with three yield levels were selected based on the combination of yield level and farmer cooperation. They are: 1 low-yield farmer FL, with an average annual yield of 8.8t ha in three years -1 About; two middle-income households FM-1 and FM-2, with an average annual output of 9.5t ha in three years -1 About; one high-yield farmer, FH, with an average annual output of 10.0t·ha in three years -1 Three treatments were set up for the four representative farmers: farmer routine, regional optimization and precision rice management:

[0110] 1. FP (Farmers' Conventional Management System): This system was obtained through farmer surveys, mainly referring to farmers' planting management methods. The nitrogen fertilizer used in the experiment was urea, and the conventional nitrogen fertilizer application rate of farmers in this area was 150 kg ha -1 . Apply base fertilizer 7 days before paddy field slurry mixing, and then transplant the seedlings. The conventional management mode of farmers is adopted during the whole growth period. The conventional method is adopted for seedling cultivation, the leaf age at the transplanting period is 3.1 to 3.5 leaves, the rice planting row spacing is 30cm×14cm, and the density is 25 holes / m 2 The irrigation method adopts the irrigation mode that local farmers are accustomed to.

[0111] 2. ROM (Regional Optimization Management System): The nitrogen fertilizer used in the experiment was urea, and the urea dosage was 125 kg ha -1 . After the paddy field is stirred, the base fertilizer is applied. After the base fertilizer is applied, the rice seedlings are transplanted. Two topdressing operations are carried out on rice, respectively at the sprouting stage and the jointing stage.

[0112] 3. PRM (Precision Rice Management System): The precision management system uses the above method to carry out pre-production simulation before production, set the optimal density and nitrogen application rate, and adopt density optimization, with row spacing of 30cm×12cm and transplanting density of 27 holes / m 2 , 6 plants / hole, light dry-wet alternating irrigation technology is used throughout the growth period, optimized irrigation is adopted, side deep fertilization is used as the basal fertilizer combined with the basal fertilizer recommended by the system, conventional seedling cultivation is carried out, and the seedling age at the time of transplanting is between 3.1 and 3.5 leaves.

[0113] In the demonstration test at the farmer's field scale at the maturity stage, six 1-square-meter yield estimation areas with representative growth as much as possible were selected for each treatment in an "S" shape. The average number of ears in each area was counted and converted to calculate the number of ears per hectare. Five holes were selected continuously based on the average number of ears per hole for rice seed testing and 1000-grain weight determination. At the same time, the number of ears per unit area and the average fruit set rate of different plots were also determined. In addition, six 2m2 plots with uniform growth as much as possible were selected in an "S" shape on each farmer's test plot. 2 In the yield measurement area, manual cutting and yield measurement were carried out, and the moisture content of rice grains was measured immediately after weighing after mechanical threshing, and then adjusted to the standard moisture content of 14%. The results are as follows:

[0114] 1. Rice yield under different treatments

[0115] The theoretical yield of low-yield farmers treated with FM is 8.85 t·ha -1 The actual field yield is 8.19 t·ha -1 In comparison, the theoretical and measured yields of ROM are 9.73 and 9.37 t·ha, respectively. -1 , and PRM were 9.86 and 9.48 t·ha -1 Compared with ROM and PRM, the low yield of farmers is mainly due to the fact that the number of ears per hectare is 4.73×10 6 ha -1 , which is 0.38×10 less than ROM and PRM respectively. 6 ha -1 , 0.44×10 6 ha -1 Relatively speaking, the differences in the number of grains per ear, fruit setting rate, and thousand-grain weight were not obvious. ROM and PRM mainly increased the number of ears per hectare for low-yield farmers. In addition, the harvest index of ROM and PRM was higher than that of FM, which were 52.2%, 52.64%, and 47.22% respectively. The harvest index of low-yield farmers was lower.

[0116] In two middle-income households, the measured yields of the FM treatment were 9.17 t·ha -1 and 9.05t·ha -1, which is significantly lower than the yields of ROM and PRM, with ROM being 9.37 t·ha -1 , 9.24t·ha -1 , PRM is 9.87t·ha -1 、9.56t·ha -1 . Compared with FM, the yield of both ROM and PRM increased by 2%, and compared with FM, the yield of the two middle-income farmers increased by 8% and 6% respectively. The yield of the two middle-income farmers was generally higher than that of the low-yield farmers. The first middle-income farmer had a higher number of grains per ear than the low-yield farmer, and the second middle-income farmer had a higher fruit set rate than the low-yield farmer. The first middle-income farmer, ROM and PRM, mainly increased the number of ears per hectare and the number of grains per ear, thus obtaining a higher yield than FM. The second middle-income farmer mainly increased the number of ears per hectare to obtain a higher yield. Among the middle-income level farmers, the harvest index ranged from 52.15 to 55.62%. The harvest index of ROM and PRM was generally higher than that of FM, but the difference in harvest index between farmers at this level was small.

[0117] Among high-yield farmers, the overall yield level was high, with the measured yields of FM, ROM and PRM being 9.47 t·ha respectively. -1 、10.01t·ha -1 and 10.17t·ha -1 , increasing yield by 5% and 7% respectively, which is the smallest increase compared with other types of farmers. ROM and FM mainly increased the number of ears per hectare, thus slightly increasing the yield. The overall yield level of high-yield farmers is higher than that of low-yield and middle-yield farmers, and the thousand-grain weight of high-yield farmers is higher than that of low-yield and middle-yield farmers; in addition, the number of ears per hectare is also higher than that of other types of farmers, and the number of grains per ear is higher than that of other types of farmers. Among high-yield farmers, the harvest index is 53.93%, 52.31%, and 52.98% respectively. Compared with FM, the harvest index of ROM and PRM is lower.

[0118] In summary, at the low-yield farmer level, PRM increased production by 16% compared to farmers and 1.2% compared to ROM; at the middle-yield farmer level, PRM increased production by 7% compared to farmers on average, for example, ROM increased production by 4%; for high-yield farmers, the yield increase effect of PRM and ROM was not obvious, increasing production by 5% and 7% respectively. Overall, PRM increased production by 10.2% compared to FM.

[0119] 2. Nitrogen partial productivity (N PFP, kg·kg -1 )

[0120] Nitrogen partial productivity (N PFP, kg·kg -1 ) = rice grain yield per unit area / nitrogen application rate per unit area

[0121] Rice nitrogen fertilizer partial productivity under various rice management systems by different farmers Fig.17 The partial nitrogen fertilizer productivity of cold-region rice was different under different farmers' yield levels and three different management strategies, and the precision management was not low in any case. There were significant differences between the FP treatments of all different types of farmers and ROM and PRM. The partial nitrogen fertilizer productivity of rice in FM was 60.5-68.7 kg kg -1 The lowest is for low-yield farmers, and the highest is for high-yield farmers. -1 and 68.7 kg kg -1 , with 8.2kg·kg -1 Compared with ROM and PRM, the PFP of PRM was higher than that of ROM in both low-yield and high-yield households, which were 72.1 kg·kg -1 and 84.9 kg kg -1 , but it did not reach a significant level. In the two middle-income farmers, the difference in PFP between PRM and ROM reached a significant level, which was significantly higher than that of ROM treatment, which was 86.7 kg·kg -1 and 79.7 kg kg -1 , among the middle-income 1 and middle-income 2 farmers, the nitrogen fertilizer use efficiency was increased by 14.4% and 6.3% respectively compared with ROM, and the average nitrogen fertilizer use efficiency was increased by 10.35%. In general, the PFP of the PRM management system is higher than the FP, except for the performance of low-yield farmers, and the performance of other levels of farmers is higher than ROM.

[0122] Field tests show that the three different strategies may produce different differences on different types of farmers. For low-yield farmers, PRM and ROM can effectively increase yield and nitrogen fertilizer partial productivity compared with FP. For middle-income farmers, both PRM and ROM can increase yield, but PRM can more effectively increase nitrogen fertilizer partial productivity. For high-yield farmers, the yield difference of the three management strategies is not large, but PRM and ROM can increase nitrogen fertilizer partial productivity, and PRM is the highest, which ultimately proves that PRM management can be suitable for different types of farmers. It can be seen that the rice cultivation precision management method based on the crop growth model of the present invention is feasible in actual rice planting.

[0123] It can be seen from the above embodiments that the present invention provides a cold-region rice precision cultivation evaluation method based on the CERES-Rice model. Based on the CERES-Rice model and many years of meteorological data, research on cold-region rice precision management was carried out. According to the growth and development law of Jiansanjiang cold-region rice, and to simplify the difficulty of the research, three different time periods were set, namely the early stage (May 20-June 25), which is the effective tillering stage, and also the nutritional growth stage of rice; the middle stage (June 25-July 31), which is the ear growth stage, which is part of the reproductive growth of rice; and the late stage (July 31-September 30), which is the fruiting stage. The three time periods were matched and combined with normal temperature (M), as well as extreme low temperature (Q1) and extreme high temperature (Q3), and a total of 27 groups (3 3 ) Different meteorological data types. The CERES-Rice model was used to simulate two different varieties, Longjing 31 and Longjing 21, under different climate years, with different densities and nitrogen application rates, and finally different yields, MNRs (net economic returns) and NPFPs (partial nitrogen productivity). The results show that under different meteorological conditions, the obtained yields, MNRs and NPFPs are different. Under higher temperature conditions, it is easier to obtain higher yields. With the increase of density, the yield and MNR can increase, but the rate of increase slows down until it does not change. Similarly, with the increase of nitrogen application, the yield and MNR will also increase, but the increase will decrease until it does not change. Under different meteorological conditions, the NPFP of cold-region rice decreases with the increase of nitrogen application, increases with the increase of density under low temperature conditions, and has no correlation with NPFP under normal temperature. Therefore, different nitrogen application rates and densities should be set under different meteorological conditions to affect climate changes and ultimately achieve the goal of high yield and high efficiency. In order to compare the effects of three different cold-region rice management strategies, namely, conventional treatment (FP), regional optimization treatment (ROM) and precision management pre-production planning (PRM), under different meteorological conditions, the present invention simulated three different strategies based on the real meteorological data of Jiansanjiang area for 65 years, and compared them. It was finally proved that PRM can obtain higher yield and MNR, but NPFP is slightly lower than ROM, but both are better than conventional treatment of farmers. It is theoretically proved that precision management pre-production planning (PRM) can increase production and income, and improve nitrogen fertilizer utilization efficiency.

[0124] The above is only a preferred embodiment of the present invention. It should be pointed out that for ordinary technicians in this technical field, several improvements and modifications can be made without departing from the principle of the present invention. These improvements and modifications should also be regarded as the scope of protection of the present invention.

Claims

1. A rice cultivation precision management method based on a crop growth model, characterized in that: The specific steps are as follows: Acquire historical meteorological data, and predict the meteorological data of the current season based on the historical meteorological data combined with the meteorological data that has occurred in the current season; The historical meteorological data includes temperature, and the temperature includes the daily minimum temperature and the daily maximum temperature; According to the forecasted meteorological data of the season, different nitrogen application rates and different planting densities are set, and the crop growth model is driven multiple times to obtain different yields under different nitrogen application rates and different planting densities. Construct a marginal net benefit formula for nitrogen application and seedling cost, use marginal net benefit as a basis for judgment, determine the target yield for the season, and then get the optimal nitrogen application amount and planting density based on the target yield; The marginal net benefit formula is: MNR=Yield*Price—Nrate*Ncost—Seedling*Scost; Among them, MNR is marginal net revenue, Yield is the simulated rice yield under the meteorological conditions predicted in the season, different nitrogen application rates, and different density conditions, in kg ha -1 , Price is the price of rice; Nrate is the nitrogen application rate, unit is kg·ha -1 , Ncost is the price of nitrogen fertilizer; Seedling is the transplanting or sowing density per hectare, Scost is the cost of seeds or seedlings; The specific steps for predicting the seasonal meteorological data are as follows: Determine the meteorological data for each year in history; divide the meteorological stages according to the three key growth nodes of rice, and divide each stage into normal year type, high temperature year type, and low temperature year type based on historical meteorological data. Divide the entire growth period into 27 meteorological types and determine the meteorological type to which each historical year belongs; Obtain the actual measured data of the daily maximum and minimum temperatures that have occurred in the current year, compare them with the meteorological data of historical years, and use the meteorological data type of the closest historical year as the predicted meteorological type for the current year; According to the forecast meteorological type of the year, the highest temperature of the forecast meteorological type is obtained, and then the lowest temperature is calculated according to the linear transformation equation of the highest temperature and the lowest temperature to obtain the forecast meteorological data of the season; The specific steps of the multiple driving crop growth model are: Different nitrogen application rates were set with fixed step lengths. To improve the calculation speed, 10 different nitrogen application rates were set. Set different sowing or transplanting densities with fixed step lengths. To simplify the calculation, set 10 different densities. By combining different densities and different nitrogen application rates, 100 cultivation plans were generated, which were input into the crop growth model in turn. Then, combined with the predicted meteorological data of the season, 100 corresponding yield data were obtained.

2. The rice cultivation precision management method based on crop growth model according to claim 1, characterized in that: Also includes: The application amount of phosphorus fertilizer and potash fertilizer is determined as follows: According to the results of MNR, the nitrogen application and sowing and transplanting schemes with the highest value can be selected, and the optimal target yield for the season can be determined; After the autumn harvest of the previous season, the available phosphorus and available potassium content of the soil were measured; Using the target yield of the model data and the phosphorus and potassium nutrient content of the soil, the optimal application amount of phosphorus and potassium fertilizers is obtained according to the target yield method; Finally, before production, the target yield for the season, nitrogen application amount, potassium application amount, phosphorus application amount, transplanting or sowing density can be obtained, enabling detailed planning of pre-production.

3. The rice cultivation precision management method based on crop growth model according to claim 1, characterized in that: The three key growth nodes of rice are: effective tillering stage, panicle formation stage and fruiting stage.

Images