Novel method for monitoring crop growth

Abstract

The invention belongs to the technical field of crop growth monitoring, and provides a novel crop growth monitoring method, which comprises the following steps: S1, optical PPFD quantum sensors are placed at horizontal and vertical distance fixing positions among crop rows according to a space grid method, the transverse and longitudinal intervals of the plurality of optical PPFD quantum sensors are 20cm, and the distance between the optical PPFD quantum sensors is 10cm; collecting PAR data in the canopy at a specific time frequency; s2, before use, calibrating the sensor according to an in-situ calibration method; s3, an optical PPFD quantum sensor is adopted to measure the PAR level of the crop in the growth stage; s4, preprocessing the original FIPAR data, removing error entries, and calculating a missing value by using an inverse distance weighting method; s5, performing visual modeling on the FIPAR data by adopting a space-time Kriging (STK) model, a fixed rank Kriging (FRK) model and an integral difference equation (IDE) model; according to the invention, the optical PPFD quantum sensor can be used for monitoring the spatial and temporal changes of the canopy and PAR of the crop, and the growth condition of the crop can be reflected by monitoring the spatial and temporal variability of the PAR of the canopy.

Description

Technical Field

[0001] The invention belongs to the technical field of crop growth monitoring, in particular to a new method for crop growth monitoring. Background Art

[0002] Crop growth monitoring is an important part of agricultural information, reflecting the growth status of crops. Real-time and accurate monitoring of growth can provide support for producers' decision-making and is very important for crop production safety. In the vast majority of cases, field surveys and remote sensing methods are the main tools for obtaining crop growth information. Both survey methods have their own advantages in practical use. As a traditional survey method, field surveys are simple, accurate, and widely applicable. Remote sensing provides a novel method for crop growth monitoring and demonstrates its capabilities in field-scale and large-area assessments. However, the shortcomings of these two methods are obvious. Field surveys are labor-intensive, time-consuming, and labor-intensive, and cannot be applied over large areas. The high technical content, high threshold for use, and low accuracy of remote sensing hinder the practical application of remote sensing technology. Therefore, this study aims to explore a new idea for crop growth monitoring that is both fast and accurate, and can be adapted to farm and regional scales.

[0003] Crop canopies play a vital role in yield formation. Structural properties are fundamental to canopy function, shaping resource capture and utilization, and altering the distribution and flow of water, light, and gases. Growth, development, and yield are directly related to the effective interception of solar radiation by the canopy; good canopy structure can increase crop biomass accumulation and yield. This spatiotemporal variation is particularly important in determining the light environment and subsequent photosynthetic productivity of the plant canopy. Therefore, measuring structural changes in crop canopies can reflect crop growth conditions. In previous studies, sensors were usually used to quantify structural changes because they are convenient, non-destructive, and widely used. Based on the structure of the sensor and the radiation-voltage conversion principle, it can be divided into two categories: photosynthetic photon flux density (PPFD) and PAR irradiance. PPFD sensors are based on the photoelectric effect and work by measuring the number of incident photons, regardless of the wavelength of the photons. This principle enables a wider measurement range and higher sensitivity. In addition, the immediate response to PPFD changes, determined by the time scale of natural changes, sets an upper limit on the sampling frequency. This upper limit is not limited by the technical capabilities of the sensor, allowing greater flexibility in practical applications. These advantages do not exist in PAR irradiance sensors. Therefore, PPFD sensors are more frequently used in agricultural production and technical research.

[0004] However, the relationship between light interception rates at different positions of the canopy and crop growth has not been systematically studied. Most previous studies have only measured the light interception rate at the bottom of the canopy. Current studies have found that the light interception rate of the entire canopy reaches its maximum after the flowering and boll stage and remains relatively stable for a considerable period of time until the cotton leaves decrease, which is inconsistent with the growth trend of cotton, resulting in the current method being time-consuming and the low precision of remote control technology.

[0005] Therefore, we proposed a hypothesis to quantify the spatiotemporal distribution characteristics of the fraction of intercepted photosynthetically active radiation (FIPAR) within the crop canopy using photometric sensors, aiming to provide a new monitoring method for crop growth and development. Summary of the invention

[0006] In order to solve the above technical problems, the present invention provides a new method for crop growth monitoring to solve the problems in the prior art that the light interception rate of the entire canopy reaches a maximum value after the flowering and boll stage and remains relatively stable for a considerable period of time until the cotton leaves are reduced, which is inconsistent with the growth trend of cotton.

[0007] A new method for crop growth monitoring comprises the following steps:

[0008] S1, according to the spatial grid method, optical PPFD quantum sensors are placed at fixed horizontal and vertical distances between crop rows, with a horizontal and vertical interval of 20 cm between several optical PPFD quantum sensors, and PAR data are collected in the canopy at a specific time frequency;

[0009] S2, before use, calibrate the sensor according to the in-situ calibration method;

[0010] S3, using optical PPFD quantum sensor to measure PAR level during crop growth stage;

[0011] S4, the original FIPAR data were preprocessed to remove erroneous entries and calculate missing values ​​using the inverse distance weighting method;

[0012] S5, using space-time kriging (STK), fixed-rank kriging (FRK) and integrated difference equation (IDE) models to model and predict FIPAR data;

[0013] S6, beta growth function was used to fit the spatial variation of crop growth traits (PH, AGB and LAI) and canopy under different FIPAR values, and the Pearson correlation coefficient between crop growth parameters and FIPAR was calculated.

[0014] Preferably, in S1, an optical PPFD quantum sensor is used to measure the spatiotemporal distribution characteristics of intercepted photosynthetically active radiation (FIPAR) in the crop canopy to monitor the growth status of the crop.

[0015] Preferably, in step S2, the specific steps of calibrating the sensor according to the in-situ calibration method are:

[0016] S2a, preheating the optical PPFD quantum sensor to a stable working state;

[0017] S2b, placing the optical PPFD quantum sensor under completely light-shielded conditions and measuring the dark current;

[0018] S2c, using a standard light source to illuminate the optical PPFD quantum sensor, and recording the response value of the optical PPFD quantum sensor under a known light intensity;

[0019] S2d, compares the optical PPFD quantum sensor reading with the known radiation intensity of the standard light source to determine if there is a deviation;

[0020] S2e, adjusting the optical PPFD quantum sensor parameters according to the deviation;

[0021] S2f, re-measure the response value under the standard light source to verify whether the adjusted optical PPFD quantum sensor is consistent with the known radiation intensity.

[0022] Preferably, in S3, the PAR level includes transmitted PAR and total PAR.

[0023] Preferably, in S4, data exceeding a maximum linear deviation of 1% is identified as an error entry.

[0024] Preferably, in S5, the best fit modeling method is evaluated based on the principles of standard cross validation (SCV), predictive cross validation (PCV), continuous ranked probability score (CPRS), Basis, Akaike Information Criterion (AIC) and Bayesian Information Criterion (BIC).

[0025] Preferably, in S6, the growth traits (PH, AGB and LAI) in the growing season are fitted with a β growth function, the formula of which is: Where G is the value of the unit day (t) after emergence; Gmax is the maximum value of the growth trait during the entire growing season; Te(d) is the time when the maximum value of the growth trait appears, and tm(d) is the time when the maximum rate is reached.

[0026] Preferably, in S6, the Pearson correlation coefficient is used to evaluate the correlation between the crop growth parameter and FIPAR.

[0027] Compared with the prior art, the present invention has the following beneficial effects:

[0028] 1. The present invention can monitor the temporal and spatial changes of crop canopy and PAR through optical PPFD quantum sensors.

[0029] 2. The present invention can reflect the crop growth status by monitoring the spatiotemporal variability of canopy PAR.

[0030] 3. The present invention can monitor crop growth non-destructively, quickly and in real time, solving the problems of the existing methods such as long time consumption and low accuracy. BRIEF DESCRIPTION OF THE DRAWINGS

[0031] Figure 1 This is a test diagram of the effectiveness of the canopy PAR gridding method of the present invention;

[0032] Figure 2 It is a spatiotemporal variation diagram of FIPAR of cotton canopy during the whole growing season of the present invention;

[0033] Figure 3 The predicted values ​​of canopy light interception by the IDE, FRK and STK models of the present invention in 2021 and 2020 were compared with the measured values ​​at 56, 75 and 100 days after emergence, respectively;

[0034] Figure 4 This is a three-dimensional visualization of cotton canopy FIPAR reconstructed using the IDE model;

[0035] Figure 5 It is the fitting graph of different FIPAR values ​​of the canopy over time and the fitting β growth function graph of growth traits (PH, AGB, LAI);

[0036] Figure 6 This is a graph showing the relationship between different FIPAR values ​​and growth traits (PH, AGB, LAI). DETAILED DESCRIPTION

[0037] The following embodiments of the present invention are described in further detail in conjunction with the accompanying drawings and examples. The following examples are used to illustrate the present invention, but are not intended to limit the scope of the present invention.

[0038] As attached Figure 1 To Attachment Figure 6 As shown:

[0039] Embodiment: The present invention provides a new method for monitoring crop growth, comprising the following steps:

[0040] S1, according to the spatial grid method, optical PPFD quantum sensors are placed at fixed horizontal and vertical distances between crop rows, with a horizontal and vertical interval of 20 cm between several optical PPFD quantum sensors, and PAR data are collected in the canopy at a specific time frequency;

[0041] S2, before use, calibrate the sensor according to the in-situ calibration method;

[0042] S3, using an optical PPFD quantum sensor to measure the PAR level of crops during their growth stages, where the model of the optical PPFD quantum sensor is LI-191SA;

[0043] S4, the original FIPAR data were preprocessed to remove erroneous entries and calculate missing values ​​using the inverse distance weighting method;

[0044] S5, using space-time kriging (STK), fixed-rank kriging (FRK) and integrated difference equation (IDE) models to model and predict FIPAR data;

[0045] S6, beta growth function was used to fit the spatial variation of crop growth traits (PH, AGB and LAI) and canopy under different FIPAR values, and the Pearson correlation coefficient between crop growth parameters and FIPAR was calculated.

[0046] Specifically, in S1, an optical PPFD quantum sensor is used to measure the spatiotemporal distribution characteristics of intercepted photosynthetically active radiation (FIPAR) within the crop canopy to monitor the growth status of the crop.

[0047] Furthermore, in step S2, the specific steps of calibrating the sensor according to the in-situ calibration method are:

[0048] S2a, preheating the optical PPFD quantum sensor to a stable working state;

[0049] S2b, placing the optical PPFD quantum sensor under completely light-shielded conditions and measuring the dark current;

[0050] S2c, using a standard light source to illuminate the optical PPFD quantum sensor, and recording the response value of the optical PPFD quantum sensor under a known light intensity;

[0051] S2d, compares the optical PPFD quantum sensor reading with the known radiation intensity of the standard light source to determine if there is a deviation;

[0052] S2e, adjusting the optical PPFD quantum sensor parameters according to the deviation;

[0053] S2f, re-measure the response value under the standard light source to verify whether the adjusted optical PPFD quantum sensor is consistent with the known radiation intensity.

[0054] Further, in S3, the PAR level includes the transmitted PAR and the total PAR.

[0055] Further, in S4, data exceeding the maximum linear deviation of 1% is identified as an error entry.

[0056] Furthermore, in S5, the best-fit modeling method was evaluated based on the principles of standard cross validation (SCV), predictive cross validation (PCV), continuous ranked probability score (CPRS), Basis, Akaike information criterion (AIC), and Bayesian information criterion (BIC).

[0057] Furthermore, in S6, the growth traits (PH, AGB and LAI) during the growing season were fitted with the β growth function, whose formula is: Where G is the value of the unit day (t) after emergence; Gmax is the maximum value of the growth trait during the entire growing season; Te(d) is the time when the maximum value of the growth trait appears, and tm(d) is the time when the maximum rate is reached.

[0058] Furthermore, in S6, Pearson correlation coefficient was used to evaluate the correlation between crop growth parameters and FIPAR.

[0059] From the above, we can see that:

[0060] Taking cotton as an example, the experiment was conducted at the Experimental Base of the Cotton Research Institute of the Chinese Academy of Agricultural Sciences in Anyang, Henan Province, China (36°06′N, 114°21′E). The experimental area has a warm temperate continental monsoon climate, with an annual temperature of 22°C, an annual precipitation of 318mm, and an average sunshine duration of 1473 hours during the crop growing season (April-October). The field soil is alluvial soil with a chemical property of pH 8.10 and an organic matter content of 16.32g·kg -1 , total nitrogen content 1.16g·kg -1 , effective phosphorus content 12.15 mg kg -1 , effective potassium content is 162.10 mg·kg.

[0061] 1. Experimental Design

[0062] The cotton monoculture (CM) was designed in 2020-2021 with three replicates. The cotton plant density was 8.7 plants / m-2, the row spacing was 80 cm, and the intra-row spacing was 14.4 cm. The planting specifications were all recommended by the local area. The cotton monoculture treatment was applied with 225 kg / hm -2 N, 150kg·hm -2 P2O5 and 225kg·hm -2 K2O. The rest of the field management work was carried out according to local practices.

[0063] 2. Data Collection

[0064] Optical PPFD quantum sensors are used to measure the PAR level of crops during their growth stages, including transmitted PAR and total PAR. According to the spatial grid method, optical PPFD quantum sensors are placed at fixed horizontal and vertical distances between crop rows. The horizontal and vertical intervals of several optical PPFD quantum sensors are 20 cm, and PAR is collected in the canopy at a specific time frequency. Before use, the sensor is calibrated according to the in-situ calibration method. FIPAR is calculated according to the formula.

[0065] In the present invention, FIPAR is calculated by the transmission PAR (TPAR), reflection PAR (RPAR) and total PAR (TOPAR) of the light sensor, and the formula is:

[0066]

[0067]

[0068] FIPRA=1-tPAR-rPAR#(3)

[0069] In addition, rPAR is less than 5% throughout the growing season. Therefore, FIPAR is also calculated according to the formula FIPRA = 1-tPAR.

[0070] The following table shows the sampling dates of cotton canopy PAR, agronomic traits and cotton phenological periods in 2020 and 2021

[0071]

[0072] Among them, agronomic traits include plant height (PH), aboveground biomass accumulation (AGB), and leaf area index (LAI). Due to weather restrictions, the frequency of canopy PAR collection in 2020 was relatively lower than that in 2021, but this difference had no significant effect on the final results.

[0073] 3. Data processing

[0074] During the field measurement, due to the influence of objective factors such as weather, FIPAR shows irregular and sparse spatiotemporal characteristics. The original FIPAR data is preprocessed by removing the erroneous entries, and the data exceeding the maximum linear offset of 1% is identified as an erroneous value. Then, the inverse distance weighted method is used to calculate the missing values.

[0075] Spatiotemporal kriging (STK), fixed-rank kriging (FRK) and integrated difference equation (IDE) models were used to model and predict FIPAR data, and the best-fit modeling method was evaluated based on the principles of standard cross validation (SCV), predictive cross validation (PCV), continuous ranked probability score (CPRS), Basis, Akaike information criterion (AIC) and Bayesian information criterion (BIC).

[0076] The beta growth function was used to fit the spatial variation of crop growth traits (PH, AGB and LAI) and canopy under different FIPAR values, and the Pearson correlation coefficient between crop growth parameters and FIPAR was calculated.

[0077] The growth traits (PH, AGB and LAI) during the growing season were fitted with the β growth function, and the formula was:

[0078]

[0079] Where G is the value of the unit day (t) after emergence; Gmax is the maximum value of the growth trait during the entire growing season. Te(d) is the time of maximum growth trait occurrence, and tm(d) is the time to reach the maximum rate. These parameters were estimated using the “SSbgf” function in R software (version 4.2.1). In addition, the spatial variation of different FIPAR in the cotton canopy was also fitted using the β growth function.

[0080] 4. Results

[0081] 4.1 Evaluation of the measurement accuracy of optical PPFD quantum sensor

[0082] Using the spatial grid method, the measurement differences of single optical PPFD quantum sensors at different positions of the crop canopy measurement system established in this study under the same environment were not significant (p>0.05), and the measurement error was within the system error range of this type of sensor (the maximum linearity offset was 1%). At the same time, the measurement system established in this study has good reproducibility in determining the spatial distribution of FIPAR. In three repeated measurements at similar times, the spatial distribution of FIPAR is consistent, and the reference Figure 1 , where (a), (b), and (c) are three sampling PAR repetitions within a short period of time. (a1), (b1), and (c1) represent the variability of FIPAR at 20 cm above the ground; (a2), (b2), and (c2) represent the changes in FIPAR at 40 cm above the ground; (a3), (b3), and (c3) show the variability of FIPAR at 20 cm above the ground. There is no positive difference in the variability of FIPAR at the same height for three repetitions.

[0083] 4.2 Spatial and temporal variation of canopy FIPAR during the cotton growing season

[0084] refer to Figure 2 , the change of FIPAR in the cotton canopy in the horizontal direction is smaller than that in the vertical direction. At the same vertical height, the change of FIPAR between canopies at different horizontal distances is the smallest. On the contrary, at the same horizontal distance, FIPAR is linearly negatively correlated with the increase of vertical height ( Figure 2In addition, the maximum FIPAR of cotton during the entire growth period was less than 0.8 in both horizontal and vertical directions. At different spatial locations, the light interception rate of the cotton canopy changed with time, but the fitting curve showed that the change pattern of the light interception rate of the canopy with time was close to the crop growth curve, such as the logistic growth curve ( Figure 2 (Part c).

[0085] Among them, a) and b) are the spatial variation diagrams of FIPAR measurements of cotton canopy, where a) shows the horizontal variation of the mean FIPAR of photometric sensors at different spatial locations throughout the breeding period; b) shows the vertical variation of the mean FIPAR of photometric sensors at different locations throughout the breeding period; c) shows the FIPAR variation at different locations of the cotton canopy profile. The sensors numbered 1, 2, 3 to 30 in the photometric sensor rack (from left to right, from bottom to top) vary from left to right, from top to bottom, and the red fitting curve is calculated using the least squares method.

[0086] 4.3 Spatiotemporal interpolation of light interception rate variation data of cotton canopy

[0087] By setting the canopy PAR observation frequency to 7 days, the spatiotemporal variation characteristics of cotton canopy FIPAR were obtained. Therefore, there is a partial loss of FIPAR information in both spatial and temporal scales, which hinders the ability to fully understand the spatiotemporal variation. In addition, the spatiotemporal variation characteristics of FIPAR within the cotton canopy have not been fully captured. Therefore, in order to establish a more accurate spatiotemporal model of FIPAR, we designed two types of training sets. The first type (I) involves removing the training set of FIPAR data at specific time points: FIPAR data measured at 51, 81, and 116 days after emergence in 2020 were omitted, while FIPAR data measured at 56, 75, and 100 days after emergence were excluded in 2021. The second type (II) involves randomly eliminating a total of 30 data entries distributed at different spatial locations and time points. Over the two-year time span, the results all show that among the three models, the IDE model has the closest deviation to 0, the smallest PCV, the closest SCV to 1, and the smallest CRPS in the I and II training datasets. The AIC and BIC values ​​of the IDE model were also the lowest among the three models, indicating that the IDE model was the best method to simulate the spatiotemporal light interception rate of cotton canopy in this study.

[0088] The following table shows the variance analysis of FIPAR measurements of sensors at different locations

[0089]

[0090] The predictions of the three models were then validated, and the results showed that the FIPAR values ​​predicted by the IDE model were closest to the actual measured values. This consistency was demonstrated by a strong linear relationship (R2 = 0.940 in 2020 and R2 = 0.749 in 2021) and a significant strong positive correlation (p < 0.001), see reference Figure 5 .

[0091] Figure 5 Comparison of the predicted values ​​and measured values ​​of canopy light interception by IDE, FRK and STK models at 56, 75 and 100 days after emergence in 2021 and 2020, respectively. The fitting curves are least squares curves. a) and d), b) and e), c) and f) show the predicted values ​​and measured values ​​of IDE, FRK and STK models, respectively.

[0092] 4.4 Simulation of canopy light interception and growth traits over time

[0093] Fitting different FIPAR values ​​(ranging from 0.1 to 0.9, with a resolution of 0.1) to the spatial variation of days after emergence revealed consistency with the β growth function. The goodness of fit first increased and then decreased with increasing FIPAR. 2 The values ​​were 0.6470 and 0.8750 (FIPAR = 0.9), and the maximum R 2 The values ​​are 0.9252 and 0.9877 (FIPAR = 0.1), respectively. At the same time, the beta growth function is fitted to PH, AGB, and LAI, and it is found that the R 2 The values ​​are 0.9865, 0.9953 and 0.9978 respectively, and the RSME values ​​are 3.69, 277.04 and 0.06 respectively, which are well fitted. 2 The corresponding values ​​are 7.14, 139.64 and 0.12, respectively. These results show that the β growth function effectively reflects the growth and development process of cotton.

[0094] The following table compares the effectiveness of IDE, FRK and STK models in simulating light interception in cotton canopies

[0095]

[0096] Among them, the FRK, IDE and STK models fitting cotton canopy FIPAR were cross-validated diagnostics. I: missing at a certain time point; II: missing at random. The IDE model performed better among all the diagnostics considered here, namely the bias close to 0 (the closer to 0, the better), the lowest predictive cross-validation metric value (PCV, the lower the better), the lowest standardized cross-validation metric value close to 1 (SCV, the closer to 1, the better) and the lowest continuous ranked probability score (CRPS, the lower the better). According to AIC and BIC, the lower the value, the better the model quality.

[0097] Figure 4 This is a 3D visualization of cotton canopy FIPAR reconstructed using the IDE model. On the left, the reinterpolated data was imported into Voxler software for reproduction, showing the changes in FIPAR within the canopy cross section during cotton growth. On the right, it shows the changes in different FIPAR within the canopy on spatial and temporal scales, with a, b, c, and d representing the cases when FIPAR is equal to 0.2, 0.4, 0.6, and 0.8, respectively.

[0098] Figure 5 Figure 1 is the fitting diagram of different FIPAR values ​​in the canopy over time and the fitting β growth function diagram of growth traits (PH, AGB, LAI), where a) represents the fitting of different FIPAR values ​​in the canopy over time (specific FIPAR fitting equations, parameters and R2 are shown in Supplementary Table S2), b) represents the fitting of cotton plant height over time, c) represents the fitting of cotton aboveground biomass accumulation over time, and d) represents the fitting of cotton leaf area index over time. In this study, all were fitted using the β growth function.

[0099] 4.5 Correlation between the spatiotemporal distribution of canopy light reception and agronomic traits

[0100] In this study, the Pearson correlation coefficients between different FIPAR values ​​and growth traits were calculated ( Figure 6 a). PH, AGB, and LAI were significantly positively correlated with each FIPAR value. Specifically, the highest correlation with PH was found when FIPAR was 0.2 and 0.3, while the highest correlation with AGB was found when FIPAR was 0.5 and 0.6. For LAI, the strongest correlation was found when FIPAR was 0.4 and 0.5. In addition, regression analysis of the spatial locations with the highest correlation between FIPAR values ​​and growth traits revealed that all relationships were consistent with the first-order linear fitting equation ( Figure 6bd). When the PH and FIPAR values ​​were 0.2 and 0.3, the linear fits had R2 values ​​of 0.9762 and 0.9645, respectively. When the AGB and FIPAR values ​​were 0.5 and 0.6, the corresponding R2 values ​​were 0.9509 and 0.9624, respectively. When the LAI and FIPAR values ​​were 0.4 and 0.5, the linear regressions had R2 values ​​of 0.9866 and 0.9941, respectively. All fits were significantly correlated (p<0.01).

[0101] Figure 6 Figure 1 shows the relationship between different FIPAR values ​​and growth traits (PH, AGB, LAI), where a): Pearson correlation heat map of different FIPAR changes in the canopy and growth traits (PH, AGB, PH). b), c) and d): Linear regression between the spatial structure of the canopy quantified by the photometric sensor and plant height (PH), aboveground biomass accumulation (AGB), and LAI. The spatial variation of the canopy refers to the vertical height of different FIPAR in the canopy throughout the growing season. For example, when the FIPAR value is 0.1, the FIPAR value of the canopy structure appears dynamically on the vertical trajectory as the number of days appears.

[0102] 5. Conclusion

[0103] The spatiotemporal variation of cotton canopy FIPAR was quantitatively analyzed using a photometric sensor system based on a spatial grid method and a spatiotemporal model. In addition, the potential of using FIPAR changes to characterize crops and their development was explored. The results show that the photometric sensor system based on the spatial grid method can effectively monitor the spatiotemporal variation of cotton canopy structure in real time. At the same time, after calibration verification, the IDE model can accurately simulate the dynamic variation of FIPAR in canopy structure, and its predicted value is closest to the measured value (R2=0.749-0.940). The correlation with PH is strongest when the FIPAR values ​​are 0.2 and 0.3, while the correlation with AGB reaches a peak when the FIPAR values ​​are 0.5 and 0.6. For LAI, the correlation is strongest when the FIPAR values ​​are 0.4 and 0.5. These results show that quantifying the spatial variation of FIPAR in cotton canopy can be used as a new method to evaluate crop growth. The present invention provides a theoretical basis for real-time monitoring of the spatiotemporal variation of crop canopy FIPAR, and proposes a new idea for monitoring crop growth using photometric sensors, which has important practical application significance for agricultural intelligent monitoring and crop management.

[0104] The embodiments of the present invention are provided for the purpose of illustration and description. Although the embodiments of the present invention have been shown and described above, it can be understood that the above embodiments are exemplary and cannot be understood as limitations of the present invention. Ordinary technicians in this field can change, modify, replace and modify the above embodiments within the scope of the present invention.

Claims

1. A new method for crop growth monitoring, characterized in that: The following steps are involved: S1, according to the spatial grid method, optical PPFD quantum sensors are placed at fixed horizontal and vertical distances between crop rows, with several optical PPFD quantum sensors spaced 20 cm apart horizontally and vertically, and PAR data are collected in the canopy at a specific time frequency; S2, before use, calibrate the sensor according to the in-situ calibration method; S3, using optical PPFD quantum sensor to measure PAR level during crop growth stage; S4, the original FIPAR data were preprocessed to remove erroneous entries and calculate missing values ​​using the inverse distance weighting method; S5, visual modeling of FIPAR data using space-time kriging (STK), fixed-rank kriging (FRK) and integral difference equation (IDE) models; S6, beta growth function was used to fit the spatial variation of crop growth traits (PH, AGB and LAI) and canopy under different FIPAR values, and the Pearson correlation coefficient between crop growth parameters and FIPAR was calculated.

2. A new method for monitoring crop growth as claimed in claim 1, characterized in that: In S1, an optical PPFD quantum sensor is used to measure the spatiotemporal distribution characteristics of intercepted photosynthetically active radiation (FIPAR) within the crop canopy to monitor the growth status of crops.

3. A new method for monitoring crop growth as claimed in claim 1, characterized in that: In step S2, the specific steps of calibrating the sensor according to the in-situ calibration method are: S2a, preheating the optical PPFD quantum sensor to a stable working state; S2b, placing the optical PPFD quantum sensor under completely light-shielded conditions and measuring the dark current; S2c, using a standard light source to illuminate the optical PPFD quantum sensor, and recording the response value of the optical PPFD quantum sensor under a known light intensity; S2d, compares the optical PPFD quantum sensor reading with the known radiation intensity of the standard light source to determine if there is a deviation; S2e, adjusting the optical PPFD quantum sensor parameters according to the deviation; S2f, re-measure the response value under the standard light source to verify whether the adjusted optical PPFD quantum sensor is consistent with the known radiation intensity.

4. A new method for monitoring crop growth as claimed in claim 1, characterized in that: In S3, the PAR level includes transmitted PAR and total PAR.

5. A new method for monitoring crop growth as claimed in claim 1, characterized in that: In S4, data exceeding a maximum deviation of 1% from linearity is identified as an error entry.

6. A new method for monitoring crop growth as claimed in claim 1, characterized in that: In S5, the best-fit modeling methods were evaluated based on the principles of standard cross validation (SCV), predictive cross validation (PCV), continuous ranked probability score (CPRS), Basis, Akaike information criterion (AIC), and Bayesian information criterion (BIC).

7. A new method for monitoring crop growth as claimed in claim 1, characterized in that: In S6, the growth traits (PH, AGB and LAI) during the growing season were fitted with the β growth function, whose formula is: Among them, G is the value of the unit day (t) after emergence; Gmax is the maximum value of the growth trait in the entire growing season, Te(d) is the time when the maximum value of the growth trait appears, and tm(d) is the time to reach the maximum rate.

8. A new method for monitoring crop growth as claimed in claim 1, characterized in that: In S6, Pearson correlation coefficient was used to evaluate the correlation between crop growth parameters and FIPAR.

Images