Potato leaf water content monitoring method based on crown temperature difference

Abstract

The invention discloses a potato leaf water content monitoring method based on crown temperature difference, which comprises the following steps: acquiring data by adopting an infrared thermal imager, determining a monitoring time period, and calculating to obtain the crown temperature difference of potatoes; respectively constructing a lower base line and an upper base line related to steam pressure deficit, and forming water stress indexes in different growth periods; and respectively constructing a quantitative relationship between the water stress index and the leaf and a quantitative relationship between the water stress index and the plant water content, verifying, and monitoring the water content of the potato leaf. The problems that in the prior art, a multi-growth-period dynamic monitoring and systematic threshold system is lacked, and accurate support cannot be provided for irrigation are solved, the potato leaf water content and the plant water deficit condition can be accurately indicated according to field verification CWSI, a basis is provided for precise irrigation of potatoes, and the method is suitable for large-scale popularization and application. Important technical support can be provided for sustainable production of potatoes in arid and semi-arid regions.

Description

Technical Field

[0001] This invention relates to a method for monitoring the water content of potato leaves, specifically a method for monitoring the water content of potato leaves based on the crown temperature difference. Background Technology

[0002] Real-time and accurate water deficit diagnosis is crucial for achieving water conservation and efficient water use. Crop water diagnosis and recommended irrigation include two approaches: one is irrigation amount recommendation based on soil moisture detection, and the other is irrigation amount recommendation based on the crop itself. Soil moisture detection methods include gravimetric analysis, neutron moisture detection, and time domain reflectometry (TDR). Gravimetric analysis requires multiple soil samples taken in the field and then brought back to the laboratory for moisture content testing. While accurate, this method is time-consuming, labor-intensive, and suffers from poor representativeness of single-point sampling, resulting in poor timeliness for irrigation guidance. Neutron moisture detection can repeatedly measure soil moisture content at any depth in the field without sampling, offering fast measurement speed, high accuracy, and direct reading of moisture content values. However, it can cause radiation exposure to the tester. Time domain reflectometry enables rapid soil moisture measurement with reliable results and ease of operation. However, in some sandy soils, the measuring tube may not make good contact with the soil, leading to difficulties in tube placement and poor repeatability of test results.

[0003] Canopy temperature difference (CTD) is one of the most sensitive indicators for plant diagnosis and drought stress. CTD-based water assessment is an important method for precision irrigation of crops. Monitoring CTD using infrared thermal imaging technology can serve as an effective means of evaluating crop drought tolerance, and there has been considerable research on other crops. However, potato growth and development characteristics differ significantly from other crops, and the response of CTD to water stress is still unclear. Furthermore, research on water stress indices based on CTD is lacking. Infrared thermometry has been widely applied to monitoring crop canopy temperature, becoming a common method for crop temperature detection and water assessment. Therefore, conducting relevant research on potatoes and establishing a diagnostic index system is of significant strategic importance for the sustainable use of regional water resources and sustainable agricultural development, with broad application prospects.

[0004] Existing literature 1 (Zhang Yanhong et al. Optimization of Potato Water Stress Index. China Agricultural Science and Technology Guide, 2025, 27 (3): 123-131, VIP Journal, Document ID: 7112664391, URL: http: / / dianda.cqvip.com / Qikan / Article / Detail?id=7112664391) constructed an empirical model of potato CWSI through pot experiments, determined the upper and lower baselines of CWSI, and clarified the significant negative correlation between CWSI and soil moisture content. It provided a model construction idea and parameter reference for the application of CWSI in potato water monitoring. However, this study was mainly limited to the pot environment and lacked validation under field conditions. It also did not involve the differences in model applicability between different growth stages and different varieties. Reference 2 (Chinese invention patent application with publication number CN121113178A) discloses an irrigation system based on potato canopy temperature acquisition and water stress index. This scheme uses infrared thermal imaging technology to acquire canopy temperature and combines it with the CWSI index to achieve automatic irrigation decisions. However, its technical solution mainly focuses on the construction of the hardware system. It lacks in-depth research and quantitative standards on the determination of the threshold of the CWSI model at specific potato growth stages, and the precise response relationship under different water stress levels. The applicability and reliability of the system need further verification. Reference 3 (Jia L, et al. Leaf-Air temperaturedifference as a reliable indicator for potato water status. Frontiers in Plant Science, 2025, 16: The study (1609350, URL: https: / / www.frontiersin.org / journals / plant-science / articles / 10.3389 / fpls.2025.1609350 / abstract) confirmed that the leaf-temperature difference (LAD) of potato is significantly correlated with irrigation amount, plant water content, and soil moisture. It identified the fourth leaf as the optimal monitoring site, established a binomial regression model between LAD and yield, and derived a yield threshold. This is a core foundational study for monitoring potato water and guiding irrigation based on canopy-temperature difference. However, this study only focused on temperature monitoring of a single leaf, failing to fully consider the spatial heterogeneity of the overall temperature distribution in the potato canopy. Furthermore, the stability and universality of the binomial regression model under different environmental conditions require more experimental data. Additionally, this study did not effectively integrate the LAD index with the existing CWSI model, and the complementarity and synergistic application mechanism between the two indices remains unclear.In summary, although existing technologies have made some progress in the diagnosis of potato crown temperature difference and moisture, they still have significant shortcomings in areas such as model verification in field environments, dynamic monitoring of multiple growth stages, integrated application of multiple indicators, and construction of a systematic irrigation decision threshold system. These shortcomings make it difficult to directly meet the actual production needs of precision irrigation for potatoes in the arid and semi-arid regions of Inner Mongolia. Summary of the Invention

[0005] The purpose of this invention is to provide a method for monitoring potato leaf water content based on crown temperature difference (CWSI). This method solves the problem that existing technologies lack dynamic monitoring of multiple growth stages and a systematic threshold system and field validation, which cannot provide accurate support for irrigation. According to field validation, CWSI can accurately indicate potato leaf water content and plant water deficit, enabling precision irrigation of potatoes and providing important technical support for sustainable potato production in arid and semi-arid regions.

[0006] To achieve the above objectives, this invention provides a method for monitoring the water content of potato leaves based on crown-temperature difference, the method comprising: Step 1: Use an infrared thermal imager to acquire data, determine the monitoring period, and calculate the crown temperature difference of the potato. Step 2: Based on the monitoring period and crown temperature difference, lower and upper baselines related to vapor pressure deficit were constructed during the potato tuber formation and expansion stages, respectively, to form water stress indices for different growth stages; Step 3: Construct and verify the quantitative relationships between water stress index and leaf moisture content and between water stress index and plant water content, respectively, for monitoring potato leaf moisture content.

[0007] Preferably, in step one, the infrared thermal image has a wavelength of 8~14 μm, an infrared resolution of 384×288, a temperature measurement range of -20~550 ℃, a NETD of 40 mK, an accuracy of ±2 ℃ or ±2%, is placed 0.90 m above the canopy, and the lens collects the canopy thermal image vertically downwards, with the emissivity parameter set to 0.98.

[0008] Preferably, in step one, the monitoring period is during the period from 10:00 to 16:00 Beijing time under sunny weather conditions; continuous monitoring is carried out using an IRT sensor, which is installed on an adjustable bracket in the furrow, arranged perpendicular to the crop row, tilted downward at 45° to align with the canopy, and the probe height is kept 0.30 m above the canopy and dynamically adjusted with crop growth. The temperature signal is output as an average value at 10-minute intervals.

[0009] Preferably, in step two, the lower baseline equation for the tuber formation period is NWSB = -1.26×VPD + 0.14, and the coefficient of determination R0 2=0.88; the lower baseline equation for the tuber enlargement period is NWSB = -1.84×VPD + 1.73, with a coefficient of determination R. 2 =0.92, where NWSB is the lower baseline of the Crop Water Stress Index (CWSI) and VPD is the vapor pressure deficit.

[0010] Preferably, in step two, the upper baseline during the tuber formation period is 3.7°C, and the upper baseline during the tuber enlargement period is 3.55°C. This upper baseline is determined by monitoring the extreme values ​​of the crown temperature difference under rain-fed treatment.

[0011] Preferably, in step two, the water stress index (CWSI) is calculated as follows: CWSI = (1); In equation (1), NWSB is the lower baseline of the Crop Water Stress Index (CWSI), NTB is the upper baseline of the Crop Water Stress Index (CWSI); Tc is the canopy temperature, Ta is the air temperature at the same time, and Tc-Ta is the canopy temperature difference. The calculation of water vapor pressure deficit is as follows: (2); In equation (2), RH is the relative humidity; Tc is the canopy temperature; and Ta is the air temperature at the same time.

[0012] Preferably, in step three, the quantitative relationship between the water stress index and the leaves is verified to be as follows: during the tuber formation period, the MAE is 0.38%, RMSE is 0.50%, and NRMSE is 11.8%; during the tuber enlargement period, the MAE is 0.53%, RMSE is 0.62%, and NRMSE is 15.1%.

[0013] Preferably, in step three, the quantitative relationship between the water stress index and the leaf and the quantitative relationship between the water stress index and the plant water content are both negatively linearly correlated.

[0014] More preferably, the quantitative relationship between the water stress index and the leaves is: the coefficient of determination R during tuber formation period. 2 =0.800, CWSI = -5.312LWC + 87.49, CWSI = -7.156LWC + 88.46 (I+5, R 2 =0.882); coefficient of determination R during tuber enlargement period 2 =0.784, CWSI = -4.976LWC + 86.83, CWSI = -5.838LWC + 87.35 (I+5, R 2 =0.836), where CWSI is the water stress index and LWC is the leaf water content; for every 0.1 increase in CWSI, LWC decreases by an average of 0.50 to 0.53 percentage points.

[0015] More preferably, the quantitative relationship between the water stress index and the plant water content is: the coefficient of determination R during tuber formation period. 2 =0.778, CWSI = -5.817PWC + 89.12, CWSI = -6.906PWC + 89.56 (I+5, R 2 =0.802); coefficient of determination R during tuber enlargement period 2 =0.760, CWSI = -5.075PWC + 87.51, CWSI = -6.904PWC + 88.43 (I+5, R 2 =0.838), where CWSI is the water stress index and PWC is the plant water content; the coefficient of determination R for tuber formation and tuber enlargement stages is 0.838. 2 All are between 0.760 and 0.778.

[0016] Preferably, the real-time crop water stress index is obtained, and the plant water status is judged based on the quantitative relationship between the water stress index and the leaves or / and the quantitative relationship between the water stress index and the plant water content, so as to formulate irrigation decisions.

[0017] This invention provides a method for monitoring potato leaf water content based on crown-temperature difference, which solves the problems of existing technologies lacking field validation, dynamic monitoring across multiple growth stages, and a systematic threshold system, thus failing to provide accurate support for irrigation. It has the following advantages: This invention conducts field measurements between 10:00 and 16:00 on sunny days. Water stress indices (CWSI) are constructed during the critical tuber formation and tuber enlargement stages of potato water requirement. Based on the CWSI, the water stress status of potatoes is quantified at different time periods. The upper baseline (UL) of the potato CWSI is as follows: UL = 3.70°C during tuber formation and UL = 3.55°C during tuber enlargement. The lower baseline (NWSB) is determined by a formula. During tuber formation, the leaf water content threshold is 86.37%, corresponding to a CWSI of approximately 0.375. During tuber enlargement, the leaf water content threshold is 85.64%, corresponding to a CWSI of approximately 0.321. This invention, based on field validation, demonstrates that CWSI can accurately indicate potato leaf water content and plant water deficit, enabling precision irrigation of potatoes and providing important technical support for sustainable potato production in arid and semi-arid regions. Attached Figure Description

[0018] Figure 1 This is a graph showing the upper and lower baseline relationships of Tc-Ta and VPD during the two key growth stages of potatoes in this invention, as well as the upper threshold results for extreme drought.

[0019] Figure 2This is a diagram showing the model construction results of CWSI and leaf and plant water content during the two key growth stages of this invention.

[0020] Figure 3 This is a diagram showing the verification results of the leaf and plant water content model of CWSI during the two key growth stages of this invention.

[0021] Figure 4 This is a graph showing the dynamic changes in plant and leaf water content before and after irrigation events during key growth stages of potatoes under different irrigation treatments according to the present invention.

[0022] Figure 5 This is a schematic diagram of the linear-platform fitting of leaf water content and comprehensive score during the tuber formation and tuber enlargement stages of this invention. Detailed Implementation

[0023] The technical solutions in the embodiments of the present invention will be clearly and completely described below. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0024] Currently, water deficit is a major limiting factor for potato production in arid and semi-arid regions of northern China, and the lack of accurate water diagnosis and recommended irrigation methods severely restricts the improvement of potato yield and quality. Therefore, this invention provides a method for monitoring potato leaf water content based on crown-temperature difference (Tc-Ta), which can predict the water deficit status of potato plants in real time, quickly, and non-destructively, thereby enabling scientific irrigation and ensuring that potatoes achieve maximum yield and benefits with minimal water usage.

[0025] The research area of ​​this invention is located in Kebuer Town, Chayouzhong Banner, Ulanqab City, Inner Mongolia Autonomous Region, China (41°17´59.81"N, 122°33´26.65"E, altitude 1780 m). It belongs to the mid-temperate continental climate zone with a frost-free period of 90-110 days. The annual average temperature, sunshine hours, and total solar radiation are 1.3°C, 3010.9h, and 138.9 kcal / cm², respectively. The annual average rainfall is less than 300 mm, while the annual average evaporation is about 2000 mm. The wilting coefficient is 8.9% (volume water content), and the organic matter content is 21.7 g / kg. Meteorological data including solar radiation, wind speed, temperature, precipitation, and relative humidity were recorded. Initial soil samples were collected from 0 to 100 cm depth. The contents of sand, silt, and clay at different soil depths (0–20, 20–40, 40–60, 60–80, and 80–100 cm) were measured using a Malvern laser particle size analyzer (Mastersizer 2000). Based on the particle size analysis at different depths, the soil type in the study area was determined to be mainly sandy loam.

[0026] This invention uses Excel 2021 to organize and perform preliminary statistical analysis on the raw data. SPSS 21.0 is used for analysis of variance: one-way ANOVA is used for univariate experiments. When the main effect reaches a significant level (p<0.05), the least significant difference (LSD) method is used for post-hoc multiple comparisons to test the differences between the means of different treatments. Simultaneously, for experiments involving multiple factors, two-way and three-way ANOVA are further performed in SPSS to test for interaction effects. Data plotting is performed using OriginPro 2021.

[0027] Example 1 A method for monitoring potato leaf water content based on crown temperature difference, the method comprising: Step 1: Use an infrared thermal imager to collect thermal images of the potato canopy, determine the monitoring period, and calculate the temperature difference between the potato canopy and the air temperature.

[0028] The canopy temperature difference (Tc-Ta) of potatoes was measured using a handheld infrared thermal imager (Fluke TiS75+, USA). Fixed monitoring devices were deployed at different irrigation treatment groups (W0: rainfed; W1-W4 irrigation amounts were 75, 150, 225, and 300 mm, respectively) and the rainfed treatment (W0). Before the seedling stage, three IRT sensors (IR630D, Beijing Time Domain Technology Co., Ltd.) were randomly installed at each treatment, for a total of six sensors. Temperature signals were acquired using an i-Logger (G30P, Beijing Time Domain Technology Co., Ltd.), and average values ​​were output at 10-minute intervals. The IRT sensors were mounted on adjustable supports within the furrows, perpendicular to the crop rows and tilted downwards at a 45° angle to the canopy. The probe height was maintained approximately 0.30 m above the canopy and dynamically adjusted as the crop grew to ensure the canopy remained the primary observation target. The infrared thermal imager operates in the 8–14 μm band, with an infrared resolution of 384 × 288, a temperature measurement range of −20–550 ℃, a D:S ratio of 524:1, an IFOV of 1.91 mrad, a NETD of 40 mK (0.04 ℃), and an accuracy of ±2 ℃ or ±2%. During measurements, the infrared thermal imager is positioned approximately 0.90 m above the canopy, with the lens pointing vertically downwards to acquire thermal images of the canopy, ensuring that the field of view is primarily covered by the canopy. One typical irrigation cycle is selected for each growth period, with observations conducted 1 day before irrigation and 1, 3, and 5 days after irrigation. Observations are conducted from 07:00 to 19:00 Beijing time, with measurements taken hourly, and 5 thermal images acquired per plot. Thermal image data was processed using SmartView Classic 4.4 (Fluke): For each thermal image, a canopy area (ROI) was delineated, avoiding cell edges and areas clearly containing bare soil or shadows. The average temperature within the ROI was extracted as the canopy temperature (Tc) for that thermal image. The Tc values ​​of five thermal images taken at the same time from the same cell were averaged to obtain the Tc for that cell at that time. The thermal imager and IRT emissivity parameters were set to their default settings (ε=0.98). The atmospheric temperature at the same time was Ta, and the canopy temperature difference was obtained by subtracting Ta from Tc.

[0029] Analysis of the intraday variation of potato canopy temperature difference under different plant water conditions revealed that during the early morning (07:00-08:00) and late afternoon (19:00) when light intensity is weaker, there were no significant differences among different irrigation treatments, and the canopy temperature difference was relatively small (0.09-0.20℃). The differences between treatments increased from 09:00 and remained highly significant from 10:00 to 16:00. p<0.001), during which all statistical indicators reached their optimal state: the F value increased significantly (846.5~1026.1), the effect size was extremely high (η² = 0.75~0.82), and the coefficient of variation remained at a low level (CV = 0.41~0.53). The temperature difference between irrigation treatments not only had the greatest distinguishability but also the best data stability. From the change process, the crown temperature difference showed a single-peak curve with the increase of radiation, reaching the intraday peak (5.87℃) at 14:00 (see Table 1 for specific data).

[0030] Table 1 shows the sensitive periods for screening irrigation gradient differences based on crown-temperature difference (Tc-Ta) according to this invention.

[0031] Based on the comprehensive analysis of Table 1, considering statistical significance, effect size, and data stability, 10:00-16:00 was selected as the optimal monitoring period for diagnosing potato water stress.

[0032] Step 2: Based on the monitoring period and crown temperature difference, dynamic lower baselines and fixed upper baselines related to vapor pressure deficit were constructed during the potato tuber formation and expansion stages (potato tuber formation and expansion stages) to form water stress indices for different growth stages.

[0033] like Figure 4 The figure shows the dynamic changes in plant and leaf water content before and after irrigation events during key growth stages of potatoes under different irrigation treatments according to this invention. Figure a represents 2023; figure b represents 2024; W0: rainfed; irrigation amounts for W1-W4 were 75, 150, 225, and 300 mm, respectively; the horizontal axis represents the number of days with the irrigation date as the zero point (Day 0, dashed line), including 1 day before irrigation (-1), 1 day after irrigation (+1), and 5 days after irrigation (+5); points represent observed values, and broken lines connect the mean change trends of each treatment within the same stage; different lowercase letters indicate significant differences between different irrigation treatments at the same growth stage and on the same observation day (LSD, p<0.05); three-way ANOVA results: the main effect significance of irrigation treatment (I), growth stage (Stage), and observation day (Days), and the significance level of the interaction I×Stage×Days (ns indicates no significance; *, **) (These represent p < 0.05 and p < 0.01, respectively). Figure 4It was observed that plant water content (PWC) and leaf water content (LWC) dynamically responded to irrigation events. Both years of trials exhibited a typical "rapid recovery after irrigation – decline during the interval" pattern: irrigation significantly improved crop water status 1 day (+1 day) after irrigation, but as soil moisture was depleted, all indicators showed varying degrees of decline 5 days (+5 days) after irrigation. A clear gradient separation was formed among different treatments: rainfed (W0) and low-irrigation (W1) treatments remained in the low value range and showed the largest decline; conversely, medium-high irrigation (W2-W4) effectively mitigated water loss during the interval, with W3-W4 exhibiting excellent water retention capacity. This difference between treatments was particularly pronounced 5 days after irrigation, indicating that higher irrigation quotas not only increased the instantaneous recovery rate but also significantly enhanced the persistence of plants in maintaining a high water state during the irrigation interval. Three-way ANOVA showed that the main effects of irrigation treatment (I), growth period (S), and observation date (D) were all significant or highly significant (p<0.05 or p<0.01); however, the interaction among the three (I×S×D) was not significant, indicating that the above response pattern is universal across different growth stages. Although the absolute values ​​of the two-year data showed interannual fluctuations due to background meteorological conditions (such as rainfall and evaporation demand), the water differentiation pattern dominated by the irrigation gradient remained highly consistent with the temporal dynamics over the two years.

[0034] Two water-demand periods, tuber formation and tuber enlargement, were used to establish water deficit (W0) and adequate water supply (W4) treatments. The temperature difference between the potato crown and tuber was measured using an infrared thermal imager between 10:00 and 16:00. By regressing the relationship between Tc-Ta and vapor pressure deficit (VPD, kPa), the lower baseline (NWSB) and upper baseline / upper limit (UL) of the crop water stress index (CWSI) were established. The CWSI was calculated as follows: CWSI = (1); In equation (1), NWSB is the lower baseline of the crop water stress index (CWSI), NTB is the upper baseline of the crop water stress index (CWSI); Tc is the canopy temperature, Ta is the air temperature at the same time, and Tc-Ta is the canopy temperature difference.

[0035] Air temperature (Ta) and relative humidity (RH) are used to calculate the vapor pressure deficit (VPD, kPa). The calculation of the vapor pressure deficit is as follows: (2); In equation (2), RH is the relative humidity; Tc is the canopy temperature; and Ta is the air temperature at the same time.

[0036] To establish the lower baseline (NWSB) equation, linear regression analyses were performed during the potato tuber formation period, tuber enlargement period, and two combined periods over two years. The data used for constructing the lower baseline equation came only from the W4 treatment, specifically observational data collected between 10:00 and 16:00 under clear skies on the first day after sufficient irrigation (I+1) or after significant rainfall. Within each growth period, this invention first fitted the annual NWSB regression equations for 2023 and 2024, and then used analysis of covariance (ANCOVA) to assess the annual effect (slope: VPD × Year interaction term; intercept: the main effect of Year after centering on VPD). Under sufficient water supply conditions (W4), the Tc-Ta during tuber enlargement is more sensitive to VPD, thus necessitating the establishment of lower baselines for different growth periods, rather than using a single seasonal baseline.

[0037] The upper limit (UL) was determined based on the method of Idso et al. (1981) using canopy temperature observations under severe water stress with rain-fed treatment (W0). Under extreme drought conditions, the variation range of (Tc-Ta) is narrow and there is no obvious trend with VPD, so UL is regarded as a constant for different growth stages.

[0038] like Figure 1 As shown in the figure, the upper and lower baseline relationships of Tc-Ta and VPD during the two key growth stages of potato in this invention, as well as the upper threshold results for extreme drought, are illustrated. The upper row (ac) represents the tuber formation stage: a (2023), b (2024), c (two years combined); the lower row (df) represents the tuber enlargement stage: d (2023), e (2024), f (two years combined); the lower baseline NWSB is blue; and the upper baseline UL is red. Figure 1 It can be seen that the baseline under non-water stress (NWSB) showed a significant negative slope in both growth periods and over two years, and the goodness of fit was high (R). 2 The values ​​range from 0.87 to 0.94, indicating that under sufficient water supply conditions, as VPD increases, enhanced transpiration cooling leads to a decrease in Tc-Ta, and the lower baseline slope during tuber formation is between -1.22 and -1.30 (2023: y = -1.30x + 0.25, R). 2 = 0.88; 2024: y = -1.22x + 0.02, R 2 = 0.87; ANCOVA did not detect significant interannual differences in slope or intercept. p >0.05, combining the two years' data yields the final NWSB equation for different fertility periods: y = -1.26x + 0.14, R 2=0.88, where y represents Tc-Ta and x represents VPD), showing good interannual consistency. The absolute value of the baseline slope during the tuber enlargement period was generally larger (2023: y = -1.87x + 1.83, R 2 = 0.90; 2024: y = -1.79x + 1.61, R 2 = 0.94; ANCOVA also shows that there is no significant annual effect on either the slope or the intercept. p >0.05, merge NWSB as: y=-1.84x+1.73, R 2 =0.92, where y represents Tc-Ta and x represents VPD. It is noteworthy that the NWSB slope during tuber enlargement is significantly steeper than that during tuber formation, and the interaction term of VPD×Stage is significant ( p <0.001; Figure 1 c and Figure 1 (f), thus confirming that NWSB is invariant on an interannual scale but dependent on fertility period, and supporting the use of baselines under different fertility periods. Figure 1 c and Figure 1 In f, UL is represented by a horizontal dashed line and is defined as the average value under extreme drought conditions (Tc-Ta). The resulting UL values ​​are 3.70 °C during tuber formation and 3.55 °C during tuber enlargement, representing near-zero transpiration conditions.

[0039] Step 3: Construct and verify the quantitative relationship between water stress index and leaf area and between water stress index and plant water content.

[0040] Correlation analysis and modeling were performed between CWSI and potato plant and leaf water content. During the tuber formation and tuber enlargement stages of potatoes in 2023, leaf water content (LWC), plant water content (PWC), and CWSI were simultaneously measured and correlation analysis was conducted.

[0041] During the tuber formation and tuber enlargement stages of potatoes, plant samples were collected 1 day before irrigation and 1 and 5 days after irrigation for each irrigation cycle. Five plants of uniform growth and free from pests and diseases were randomly selected from each plot and brought back to the laboratory for testing. First, the fresh weight of the whole plant was measured and recorded as FW. Then, the plant was divided into four parts: root, stem, leaf, and tuber. The samples of each part were first blanched at 105°C for 30 min, and then dried in an 80°C oven for 48 h or until constant weight. After drying, the dry weight of each part was weighed separately, and the sum was recorded as the plant dry weight DW. At the same time, the fresh weight of the leaf samples was weighed separately and recorded as LFW; after drying to constant weight, the dry weight was weighed and recorded as LDW. The plant water content and leaf water content are calculated as follows: Plant water content PWC (%) = (FW - DW) / FW × 100%; Leaf water content LWC (%) = (LFW - LDW) / LFW × 100%.

[0042] like Figure 2 The figure shows the model construction results of CWSI and leaf and plant water content during two key growth stages of this invention. Where: a and b are leaf water content during tuber formation and tuber enlargement stages, respectively; y represents CWSI, and x represents LWC; c and d are plant water content during the corresponding growth stages, respectively; y represents CWSI, and x represents PWC; the scatter plot color represents the time from the measurement day to the irrigation day: 1 day before irrigation (I-1), 1 day after irrigation (I+1), and 5 days after irrigation (I+5); the colored dashed line represents the grouped linear fit of each time series; the black solid line represents the overall fit combining all time series; the gray shaded area represents the 95% confidence region of the overall fit. Figure 2 It can be seen that during the tuber formation and tuber enlargement stages in 2023, both leaf water content (LWC) and plant water content (PWC) of potatoes showed a stable negative linear relationship with CWSI, indicating that CWSI can effectively indicate the dynamic changes in plant water status before and after irrigation events. Overall fitting results show that CWSI can explain the main part of the variation in leaf and plant water content, and has strong characterization ability. Regarding leaf water content, during the tuber formation stage (… Figure 2 a) and tuber enlargement period ( Figure 2 The coefficient of determination (R) of b) 2 The coefficients of determination for tuber formation, R0, were 0.800 and 0.784, respectively. 2 =0.800, y= -5.312x +87.49, y= -7.156x + 88.46 (I+5, R 2 =0.882); coefficient of determination R during tuber enlargement period 2 =0.784, y= -4.976x +86.83, y= -5.838x + 87.35 (I+5, R 2=0.836), where y represents CWSI and x represents LWC), the slope of the regression equation indicates CWSI ( Figure 2 For every 0.1 increase in y, LWC (where y is the median) Figure 2 The average decrease was approximately 0.50–0.53 percentage points (x). Plant water content showed a similar pattern. Figure 2 c and Figure 2 d, coefficient of determination of tuber formation period R 2 =0.778, y= -5.817x +89.12, y= -6.906x + 89.56 (I+5, R 2 =0.802); coefficient of determination R during tuber enlargement period 2 =0.760, y= -5.075x +87.51, y= -6.904x + 88.43 (I+5, R 2 =0.838), where y represents CWSI and x represents PWC), where y represents CWSI and x represents PWC), and the total R... 2 The values ​​ranged from 0.760 to 0.778, validating the reliability of CWSI as an indicator of whole-plant moisture status. This fit revealed the dynamic modulation effect of irrigation on the CWSI-moisture content relationship. During the water consumption phase 5 days after irrigation (I+5), the goodness of fit for each index generally reached its highest (R0.05 for LWC during tuber formation). 2 The regression slope increased significantly (from 0.882), and the absolute value of the regression slope also increased significantly (tuber formation period I+5: y = -7.156x + 88.46, where y represents CWSI and x represents LWC). This indicates that as soil moisture is consumed after irrigation, the sensitivity and stability of CWSI in indicative of plant water deficiency significantly increase. Two-way ANOVA confirmed that the irrigation treatment, measurement date, and their interaction all reached a highly significant level (p<0.01), indicating that the coupling relationship between plant water status and CWSI is not static, but is jointly driven by irrigation quota and irrigation timing. In summary, these results validate the physiological indicative significance of CWSI at the leaf and whole-plant scales, providing direct evidence for subsequent irrigation.

[0043] In 2024, the same irrigation amount was used to independently validate the model. The validation results show that the water content prediction model based on CWSI exhibits good consistency in both key growth stages, with the scatter plots generally distributed along a 1:1 line. The model can accurately reproduce the changes in measured leaf and plant water content.

[0044] like Figure 3The figure shows the validation results of the leaf and plant water content model for CWSI at two key growth stages of this invention, where: a and b are the leaf water content (LWC) during tuber formation and tuber enlargement, respectively; c and d are the plant water content (PWC) at the corresponding stages; mean absolute error (MAE), root mean square error (RMSE), and normalized root mean square error (NRMSE). Figure 3 It can be seen that the water content prediction model based on CWSI showed good consistency in both key growth stages, with the scatter plots generally distributed along a 1:1 line, indicating that the model can accurately reproduce the changes in measured leaf and plant water content. The validation accuracy for leaf water content was high: during tuber formation (… Figure 3 a) MAE was 0.38%, RMSE was 0.50%, and NRMSE was 11.8%; during the tuber enlargement period ( Figure 3 (b) MAE was 0.53%, RMSE was 0.62%, and NRMSE was 15.1%. Plant water content also maintained a low error level: during the formation period ( Figure 3 c) MAE was 0.67%, RMSE was 0.66%, and NRMSE was 11.8%; during the expansion period ( Figure 3 The model's mean squared error (d) was 0.48%, mean squared error (RMSE) was 0.64%, and mean squared error (NRMSE) was 13.5%. Overall, the RMSE was less than 0.66%, indicating good predictive accuracy across different growth stages. Field validation shows that the CWSI model accurately indicates potato leaf water content and plant water deficit, enabling precision irrigation and providing crucial technical support for sustainable potato production in arid and semi-arid regions.

[0045] To balance yield formation and water use efficiency, this invention uses a comprehensive evaluation index S=0.5RY+0.5RWUE to evaluate different water treatments, where RY is the relative yield and RWUE is the relative water use efficiency. Based on experimental data from 2023 and 2024, a linear-platform relationship between leaf water content and the comprehensive score was established during the tuber formation and tuber enlargement stages, respectively. Linear-platform fitting was performed, and the inflection point corresponds to the irrigation threshold during the tuber enlargement stage. The leaf water content measured in the field was 82%, the lowest observed value, i.e., the leaf water content under the W0 treatment. For leaf water content below 82%, the fitting results using the mathematical formula were consistent with those for leaf water content between 82% and 85%.

[0046] like Figure 5 As shown, this invention presents a linear-platform fitting diagram of leaf water content and comprehensive score during the tuber formation and tuber enlargement stages, with leaf water content as the abscissa and comprehensive score as the ordinate. The scatter points in the diagram represent measured data points from different years and treatments, and the solid line represents the fitted curve. Figure 5The results show that during the tuber formation stage, the overall score increases with increasing leaf water content. Once the leaf water content reaches a certain level, it enters a plateau zone, and further increases in leaf water content no longer significantly improve the overall score. Linear-platform fitting yielded a leaf water content threshold of 86.37% for the tuber formation stage, corresponding to a CWSI of approximately 0.375. This indicates that when the leaf water content is below this threshold during tuber formation, appropriate supplemental irrigation helps to simultaneously improve yield and water use efficiency; however, once the leaf water content reaches this threshold, further increases in water supply have limited overall benefits. During the tuber enlargement stage, the overall score also shows a trend of first increasing and then stabilizing. Linear-platform fitting yielded a leaf water content threshold of 85.64% for the tuber enlargement stage, corresponding to a CWSI of approximately 0.321. This suggests that the irrigation initiation point during the tuber enlargement stage can be controlled near this threshold to balance yield formation and water conservation. In summary, irrigation thresholds differ across different growth stages, and using thresholds based on growth stage is beneficial for improving the targeting and scientific nature of irrigation decisions. The above thresholds are recommended irrigation thresholds obtained based on the experimental conditions and comprehensive evaluation indicators, providing parameter basis for the irrigation method described in this invention.

[0047] Although the present invention has been described in detail through the preferred embodiments above, it should be understood that the above description should not be considered as a limitation of the present invention. Various modifications and substitutions to the present invention will be apparent to those skilled in the art after reading the above description. Therefore, the scope of protection of the present invention should be defined by the appended claims.

Claims

1. A method for monitoring the water content of potato leaves based on crown temperature difference, characterized in that, The method includes: Step 1: Use an infrared thermal imager to acquire data, determine the monitoring period, and calculate the crown temperature difference of the potato. Step 2: Based on the monitoring period and crown temperature difference, lower and upper baselines related to vapor pressure deficit were constructed during the potato tuber formation and expansion stages, respectively, to form water stress indices for different growth stages; Step 3: Construct and verify the quantitative relationship between water stress index and leaf area and between water stress index and plant water content, and monitor the water content of potato leaves.

2. The monitoring method according to claim 1, characterized in that, In step one, the infrared thermal image has a wavelength of 8~14 μm, an infrared resolution of 384×288, a temperature measurement range of -20~550 ℃, a NETD of 40 mK, and an accuracy of ±2 ℃ or ±2%. It is placed 0.90 m above the canopy, and the lens is vertically downward to collect the thermal image of the canopy. The emissivity parameter is set to 0.

98.

3. The monitoring method according to claim 1, characterized in that, In step one, the monitoring period is during the sunny weather conditions, from 10:00 to 16:00 Beijing time. IRT sensors are used for continuous monitoring. The IRT sensors are installed on adjustable brackets in the furrows, arranged perpendicular to the crop rows, tilted downwards at 45° to align with the canopy, and the probe height is kept 0.30 m above the canopy and dynamically adjusted with crop growth. The temperature signal is output as an average value at 10-minute intervals.

4. The monitoring method according to claim 1, characterized in that, In step two, the lower baseline equation for the tuber formation period is NWSB = -1.26×VPD + 0.14, with a coefficient of determination R0. 2 =0.88; the lower baseline equation for the tuber enlargement period is NWSB = -1.84×VPD + 1.73, with a coefficient of determination R. 2 =0.92, where NWSB is the lower baseline of the Crop Water Stress Index (CWSI) and VPD is the vapor pressure deficit.

5. The monitoring method according to claim 1, characterized in that, In step two, the upper baseline during tuber formation is 3.7°C, and the upper baseline during tuber enlargement is 3.55°C. This upper baseline is determined by monitoring the extreme values ​​of the crown temperature difference under rain-fed treatment.

6. The monitoring method according to claim 1, characterized in that, In step three, the quantitative relationship between the water stress index and the leaves was verified. During the tuber formation period, the MAE was 0.38%, RMSE was 0.50%, and NRMSE was 11.8%; during the tuber enlargement period, the MAE was 0.53%, RMSE was 0.62%, and NRMSE was 15.1%.

7. The monitoring method according to claim 1, characterized in that, In step three, the quantitative relationship between the water stress index and the leaves, and the quantitative relationship between the water stress index and the plant water content, are both negatively linearly correlated.

8. The monitoring method according to claim 7, characterized in that, The quantitative relationship between the water stress index and leaves is: coefficient of determination R during tuber formation. 2 =0.800, CWSI = -5.312LWC + 87.49; coefficient of determination for tuber enlargement period R 2 =0.784, CWSI = -4.976LWC + 86.83, where CWSI is the water stress index and LWC is the leaf water content; For every 0.1 increase in CWSI, LWC decreases by an average of 0.50 to 0.53 percentage points.

9. The monitoring method according to claim 7, characterized in that, The quantitative relationship between the water stress index and plant water content is: coefficient of determination R during tuber formation period. 2 =0.778, CWSI = -5.817PWC + 89.12; coefficient of determination for tuber enlargement period R 2 =0.760, CWSI = -5.075PWC + 87.51, where CWSI is the water stress index and PWC is the plant water content; the coefficient of determination R for tuber formation and tuber enlargement stages. 2 All are between 0.760 and 0.

778.

10. The monitoring method according to claim 1, characterized in that, Obtain real-time crop water stress index, determine plant water status based on the quantitative relationship between water stress index and leaf area and / or the quantitative relationship between water stress index and plant water content, and formulate irrigation decisions.

Images