Sap flow based plant stress determination

Abstract

A system for determining plant stress includes at least one sap flow sensor obtaining sap flow measurements from at least one plant, a meteorological data source providing evapotranspiration data, and a processor that analyzes correlations between the sap flow measurements and evapotranspiration data, generates a stress indicator based on the correlations, and determines a stress condition of the plant based on the stress indicator. A corresponding method comprises obtaining sap flow measurements, receiving evapotranspiration data, analyzing correlations, generating a stress indicator, and determining a stress condition.

Description

[0001] SAP FLOW BASED PLANT STRESS DETERMINATION

[0002] CROSS-REFERENCE TO RELATED APPLICATIONS

[0003] This application claims priority to U.S. Provisional Patent Application 63 / 714,163, filed on October 31, 2024, which is hereby incorporated by reference in its entirety.

[0004] TECHNICAL FIELD

[0005] The present invention relates to plant monitoring and agricultural productivity systems, and more particularly to sap flow measurement systems and methods for determining plant stress conditions.

[0006] BACKGROUND

[0007] Effective plant cultivation requires the detection and treatment of various sources of plant stress in real-time. Plants may be influenced by a wide range of factors that impair growth and development, such as temperature extremes, water deficiencies, nutrient imbalances, diseases, and pathogens. Many stress factors can be difficult to detect and accurately identify using conventional monitoring approaches. Traditional agricultural monitoring processes primarily focus on detecting water-related issues, utilizing indirect indicators such as soil moisture levels or climate data. Other techniques involve direct measurements of plant characteristics, but these can be tedious and timeconsuming and require trained personnel. Growers often rely on visual symptoms or lagging indicators, which delay corrective actions, diminishing crop health and productivity.

[0008] Several existing tools are available for plant stress monitoring. Thermal imaging can be applied for measuring leaf temperature as an indication of stomatai conductance and transpiration changes. However, thermal imaging cannot provide detailed information about the specific cause of plant stress and may lead to false positives or misinterpretations due to environmental factors such as wind, humidity, and ambient temperature variations. Dendrometers measure variations in plant organ diameter, which may result from changes in plant water content and growth patterns. However, dendrometer measurements provide indirect indications limited to water availability aspects, while variations in plant organ diameter may be affected by factors beyond water stress, such as temperature fluctuations and growth patterns. Dendrometer readings may also lag behind immediate stress events and require frequent maintenance and recalibration. Pressure chambers measure water potential of plant tissues by applying compressed gas pressure until sap exudes from cut plant samples. While useful for understanding water tension, pressure chamber measurements provide brief snapshots and are invasive and destructive, requiring cutting plant portions. The process is time-consuming and requires expertise to perform measurements accurately. Infrared gas analyzers (IRGA) and porometers can measure rates of photosynthesis, transpiration, and stomatai conductance by analyzing gas exchange. However, these measurements are typically conducted on small scales with individual leaves or plant segments, requiring manual effort for each measurement. The results may not be representative of entire plants or plant fields and are sensitive to environmental factors.

[0009] Accordingly, existing plant stress detection tools have various limitations in providing comprehensive, real-time monitoring across different types of stress factors. There remains a need for improved methods and systems that can accurately identify a wide range of plant stress factors for enhanced agricultural monitoring and productivity.

[0010] SUMMARY

[0011] This summary is provided to introduce a selection of concepts in a simplified form that are further described below in the detailed description. This summary is not intended to identify key features or essential features of the claimed subject matter, nor is it intended to be used as an aid in determining the scope of the claimed subject matter.

[0012] According to an aspect of the present disclosure, a system for determining plant stress is provided. The system includes at least one sap flow sensor configured to obtain sap flow measurements from at least one plant. The system includes a meteorological data source configured to provide evapotranspiration data relating to the at least one plant. The system includes a processor configured to receive the sap flow measurements and the evapotranspiration data, analyze correlations between the sap flow measurements and the evapotranspiration data, generate a stress indicator based on the analyzed correlations, and determine a stress condition of the at least one plant based on the stress indicator.

[0013] According to other aspects of the present disclosure, the system may include one or more of the following features. Analyzing correlations may include determining at least one correlation coefficient between the sap flow measurements and the evapotranspiration data. The processor may be further configured to receive meteorological data and filter out sap flow measurements obtained during precipitation periods based on the meteorological data. The processor may be further configured to evaluate sensor stability of the at least one sap flow sensor and filter out sap flow measurements obtained from sap flow sensors that fail to meet predefined stability criteria. The processor may be further configured to receive meteorological data, train at least one predictive model to predict sap flow based on the meteorological data, generate a dependence analysis for the at least one predictive model to identify relationships between evapotranspiration and sap flow, identify an evapotranspiration threshold beyond which sap flow behavior becomes irregular based on the dependence analysis, and filter out sap flow measurements obtained during periods when evapotranspiration data exceeds the identified evapotranspiration threshold. The dependence analysis may include a partial dependence plot that indicates a marginal effect of evapotranspiration data on predicted sap flow. The processor may be further configured to preprocess the sap flow measurements and the evapotranspiration data by applying at least one operation selected from the group consisting of filtering, smoothing, normalization, and noise reduction. The at least one plant may include a plurality of plants, and the processor may be further configured to generate a plurality of stress indicators corresponding to the plurality of plants, aggregate the plurality of stress indicators into a group stress indicator, and apply at least one dynamic threshold to the group stress indicator based on external factors. Aggregating the plurality of stress indicators may include determining reliability metrics for individual stress indicators based on reliability of underlying sap flow measurements, and applying proportional weighting based on the reliability metrics when calculating the group stress indicator. The external factors may include at least one selected from the group consisting of time of year, seasonal variations, historical data patterns, and plant type characteristics. The at least one sap flow sensor may be configured to determine sap flow based on thermal heat transfer analysis using a heat source and a thermal sensor positioned in proximity to plant tissue. The system may further include a user device having a user interface, and a user management application configured to provide information relating to the determined stress condition through the user interface. The user management application may be further configured to provide at least one selected from the group consisting of an alert when the determined stress conditions exceeds a predefined severity threshold, a report including stress level classifications for multiple plants over a selected period of time, historical trend analysis of stress conditions, and stress mitigation recommendations based on the determined stress condition. The stress condition may include a classification selected from the group consisting of high stress, medium stress, low stress, and no stress. The processor may be further configured to determine the stress indicator at periodic intervals, and update the stress condition determination based on changing environmental conditions and new sap flow measurements.

[0014] According to another aspect of the present disclosure, a method for determining plant stress is provided. The method includes obtaining sap flow measurements from at least one plant using at least one sap flow sensor. The method includes receiving evapotranspiration data relating to the at least one plant from a meteorological data source. The method includes analyzing correlations between the sap flow measurements and the evapotranspiration data. The method includes generating a stress indicator based on the analyzed correlations. The method includes determining a stress condition of the at least one plant based on the stress indicator.

[0015] According to other aspects of the present disclosure, the method may include one or more of the following features. Analyzing correlations may include determining at least one correlation coefficient between the sap flow measurements and the evapotranspiration data. The method may further include receiving meteorological data and filtering out sap flow measurements obtained during precipitation periods based on the meteorological data. The method may further include evaluating sensor stability of the at least one sap flow sensor and filtering out sap flow measurements obtained from sap flow sensors that fail to meet predefined stability criteria. The method may further include receiving meteorological data, training at least one predictive model to predict sap flow based on the meteorological data, generating a dependence analysis for the at least one predictive model to identify relationships between evapotranspiration and sap flow, identifying an evapotranspiration threshold beyond which sap flow behavior becomes irregular based on the dependence analysis, and filtering out sap flow measurements obtained during periods when evapotranspiration data exceeds the identified evapotranspiration threshold. The dependence analysis may include a partial dependence plot that indicates a marginal effect of evapotranspiration data on predicted sap flow. The method may further include preprocessing the sap flow measurements and the evapotranspiration data by applying at least one operation selected from the group consisting of filtering, smoothing, normalization, and noise reduction. The at least one plant may include a plurality of plants, and the method may further include generating a plurality of stress indicators corresponding to the plurality of plants, aggregating the plurality of stress indicators into a group stress indicator, and applying at least one dynamic threshold to the group stress indicator based on external factors. Aggregating the plurality of stress indicators may include determining reliability metrics for individual stress indicators based on reliability of underlying sap flow measurements, and applying proportional weighting based on the reliability metrics when calculating the group stress indicator. The external factors may include at least one selected from the group consisting of time of year, seasonal variations, historical data patterns, and plant type characteristics. Obtaining sap flow measurements may include determining sap flow based on thermal heat transfer analysis using a heat source and a thermal sensor positioned in proximity to plant tissue. The method may further include providing information relating to the determined stress condition through a user interface of a user device. Providing information may include providing at least one selected from the group consisting of an alert when the determined stress conditions exceeds a predefined severity threshold, a report including stress level classifications for multiple plants over a selected period of time, historical trend analysis of stress conditions, and stress mitigation recommendations based on the determined stress condition. The stress condition may include a classification selected from the group consisting of high stress, medium stress, low stress, and no stress. The method may further include determining the stress indicator at periodic intervals, and updating the stress condition determination based on changing environmental conditions and new sap flow measurements.

[0016] According to another aspect of the present disclosure, a non-transitory computer-readable storage medium is provided. The non-transitory computer-readable storage medium contains instructions that, when executed by a processor, cause the processor to perform operations including obtaining sap flow measurements from at least one plant using at least one sap flow sensor, receiving evapotranspiration data relating to the at least one plant from a meteorological data source, analyzing correlations between the sap flow measurements and the evapotranspiration data, generating a stress indicator based on the analyzed correlations, and determining a stress condition of the at least one plant based on the stress indicator.

[0017] According to another aspect of the present disclosure, a plant monitoring apparatus for determining plant stress is provided. The plant monitoring apparatus includes a housing configured for deployment at a plant growth site. The plant monitoring apparatus includes at least one sap flow sensor mounted within the housing and configured to be positioned in contact with plant tissue to obtain sap flow measurements from at least one plant. The plant monitoring apparatus includes a communication interface configured to receive evapotranspiration data from a meteorological data source. The plant monitoring apparatus includes a processing unit operatively coupled to the at least one sap flow sensor and the communication interface, the processing unit configured to analyze correlations between the sap flow measurements and the evapotranspiration data, generate a stress indicator based on the analyzed correlations, and determine a stress condition of the at least one plant based on the stress indicator. The plant monitoring apparatus includes an output interface configured to provide information relating to the determined stress condition.

[0018] BRIEF DESCRIPTION OF THE DRAWINGS

[0019] The present disclosure will be more fully understood from the following detailed description of the embodiments thereof, taken together with the drawings in which:

[0020] FIG. 1 illustrates a network environment supporting a plant stress determination system, according to aspects of the present disclosure;

[0021] FIG. 2 illustrates a schematic diagram of information flow in the plant stress determination system of FIG. 1, according to aspects of the present disclosure;

[0022] FIG. 3 illustrates a flowchart of a method for plant stress determination, according to aspects of the present disclosure;

[0023] FIG. 4 illustrates a detailed flow diagram of data collection and ETO threshold determination phases, according to aspects of the present disclosure;

[0024] FIG. 5 illustrates a detailed flow diagram of a plant stress determination phase, according to aspects of the present disclosure;

[0025] Figures 6A-6B illustrate graphs showing comparisons of plant stress detection measurements, according to aspects of the present disclosure; and

[0026] FIG. 7 illustrates a graph depicting plant transpiration, rainfall, and stress levels over time, according to aspects of the present disclosure.

[0027] DETAILED DESCRIPTION OF EMBODIMENTS

[0028] The following description sets forth exemplary aspects of the present disclosure. It should be recognized, however, that such description is not intended as a limitation on the scope of the present disclosure. Rather, the description also encompasses combinations and modifications to those exemplary aspects described herein.

[0029] The present disclosure describes a plant stress determination system and method that utilizes sap flow measurements in combination with meteorological data analysis to accurately identify and classify various types of plant stress conditions. The system may include sap flow sensors configured to obtain real-time measurements from monitored plants, meteorological data sources providing evapotranspiration and atmospheric information, and processing modules that perform correlation analysis between sap flow and meteorological data to determine stress classifications. The method may involve establishing evapotranspiration thresholds to filter irregular behavior, implementing false positive elimination processes to exclude data collected during rainfall or unstable sensor conditions, and applying dynamic thresholds to generate reliable stress level classifications. The disclosed approach may provide continuous, non-destructive monitoring capabilities that can detect a wide range of stress factors including water deficiencies, temperature extremes, and other environmental conditions, thereby enabling enhanced agricultural productivity and plant cultivation management.

[0030] Unless otherwise defined, all terms (including technical and scientific terms) used herein have the same meaning as commonly understood by one of ordinary skill in the art to which the disclosed subject matter belongs. It will be further understood that terms, such as those defined in commonly used dictionaries, should be interpreted as having a meaning that is consistent with their meaning in the context of the specification and claims and should not be interpreted in an idealized or overly formal sense unless expressly so defined herein. Well-known functions or constructions may not be described in detail for brevity and / or clarity.

[0031] It will be understood that, although the terms first, second, etc., may be used herein to describe various elements, but these elements should not be limited by these terms. Rather, these terms are only used to distinguish one element from another element. When an element is referred to as being “on”, “attached” to, “operatively coupled” to, “connected” to, “coupled” with, “contacting”, “added to, another element, it can be directly on, attached to, connected to, operatively coupled to, operatively engaged with, coupled with, added to, and / or contacting the other element or intervening elements can also be present. When an element is referred to as being “directly contacting” another element or “directly added” to another element, there are no intervening elements present.

[0032] The term “about” or “approximately” is meant to refer to a measurable value such as an amount, a temporal duration, and the like, and encompasses variations (e.g., ±20%, ±10%, ±5%, ±1%, ±0.1%) from the specified value.

[0033] The terms “plurality” and “a plurality” include “multiple” or “two or more”. The term set when used herein may include one or more items. Unless explicitly stated, the method embodiments described herein are not constrained to a particular order or sequence. Additionally, some of the described method embodiments or elements thereof can occur or be performed simultaneously, at the same point in time, or concurrently.

[0034] Certain features of the disclosure described in the context of separate embodiments may also be provided in combination in a single embodiment. Conversely, various features of the disclosure described in the context of a single embodiment, may also be provided separately or in any suitable sub-combination.

[0035] “Disclosed embodiments”, “disclosed systems” and “disclosed methods” refer to examples of inventive ideas, concepts, and / or manifestations described herein.

[0036] This disclosure employs open-ended permissive language, indicating for example, that some embodiments “may” employ, involve, or include specific features. The use of “may” indicates that although not every embodiment may employ the specific disclosed feature, at least one embodiment employs the specific disclosed feature.

[0037] The term “repeatedly” as used herein should be broadly construed to include any one or more of: “continuously”, “periodic repetition” and “nonperiodic repetition”.

[0038] The terms “user” and “operator” are used interchangeably herein to refer to any individual person or group of persons using or operating a method or system according to an aspect of the present disclosure.

[0039] The term “plant” is used herein to refer to a eukaryotic multicellular living organism capable of photosynthesis (i.e., excluding humans and animals) and which undergoes growth and development processes (e.g., in soil or water), which may include plant features such as leaves, stems, roots, flowers, and which may (but not necessarily) produce seeds or bear edible fruits.

[0040] The term “plant growth site” as used herein refers to an area in which plants are grown and cultivated, including but not limited to: an agricultural field; an orchard; a farming pasture; a gardening plot; an enclosed structure for regulating a plant growth environment, such as a greenhouse, a bioshelter, a vertical farming plot, or a winter garden; and may include areas on land or bodies of water such as for aquatic plant growth.

[0041] The terms “plant stress” and “stress” are used interchangeably herein to broadly refer to factors or conditions that may impair or detrimentally influence plant development. Examples include biotic stresses such as plant pathogens, insects, parasites, bacteria, and fungi, as well as abiotic stresses including water deficiencies, nutrient deficiencies, hypersalinity, and temperature extremes.

[0042] FIG. 1 illustrates a network environment 100 supporting a plant stress determination system 105, according to aspects of the present disclosure. Plant stress determination system 105 includes a sensors unit 120 and at least one meteorological data source 124. System 105 is configured for determining stress of at least one plant at a plant growth site 110 having multiple plants for monitoring and cultivation. Network environment 100 includes a server 130, at least one user device 140, and a communications network 128. Server 130 includes a server processor 132 and a server database 134. Server 130 may be a cloud server associated with a cloud computing platform. User device 140 includes a user device processor 144 and a user interface 146. User device 140 may be associated with a user of system 105, such as a plant grower of a plant growth site 110. User device 140 may be embodied by any type of electronic device with computing and network communication capabilities, including but not limited to: a smartphone; a laptop computer; a mobile computer; a netbook computer; a tablet computer; or any combination of the above. Network environment 100 may include a plurality of user devices operated by multiple respective users, although a single user device 140 is depicted for exemplary purposes. Similarly, network environment 100 may include a plurality of remote servers, but a single server 130 is depicted for exemplary purposes. Communications network 128 communicatively couples various components of the plant stress determination system 105.

[0043] Sensors unit 120 is configured to obtain measurements from plants at plant growth site 110. For example, plant growth site 110 includes a first plant 112, a second plant 113, and a third plant 114, where the plants 112, 113, 114 may be of different types and characteristics. Sensors unit 120 includes at least one sap flow sensor 122 for obtaining sap flow measurements, such as sap flux density or flow rate per unit area and time. In some cases, the sap flow sensor 122 is configured to determine sap flow based on thermal heat transfer analysis using a single-needle with a heat source and a thermal sensor, as described for example in PCT International Publication WO 2022 / 234564A1 to Treetoscope Ltd, entitled “Sap flow sensor and method of determining sap flow velocity”, the disclosure of which is incorporated herein by reference. Sensors unit 120 may include multiple sap flow sensors 122, where each sap flow sensor 122 may be configured to obtain sap flow measurements of a respective plant at the plant growth site 110. For example, a first sap flow sensor 122 is respective of the first plant 112, a second sap flow sensor 122 is respective of the second plant 113, and a third sap flow sensor 122 is respective of the third plant 114. Sensors unit 110 may additionally include other types of sensors configured for obtaining additional measurements of monitored plants.

[0044] Meteorological data source 124 is operative to provide meteorological information relating to environments in which monitored plants may be located, such as plant growth site 110. Meteorological data source 124 may provide weather and climate data, such as atmospheric variables including temperature, barometric pressure, humidity, wind speed, wind direction, vapor pressure deficit (VPD), precipitation, solar radiation, and the like. Meteorological data source 124 may provide evapotranspiration (ETO) data, which reflects the transfer of water from land surfaces to the atmosphere, such as water loss from soil by evaporation and transpiration from plant leaves. The ETO data may be obtained in real-time or at periodic intervals. Meteorological data source 124 may include measured data and / or predicted data, which may be obtained from suitable monitoring instruments, such as from satellite-based remote sensing devices or a meteorological observation station. Meteorological data source 124 may include proprietary and / or publicly available data sources. In some cases, meteorological data source 124 includes an application program interface (API) such as a publicly accessible weather / climate API.

[0045] Plant stress determination system 105 may include at least some portions of server 130. Server processor 132 performs data processing and may receive instructions or information from other components of system 105. Server 130 may include a plant stress analysis module 133 operating on server processor 132. Plant stress analysis module 133 may process sensor data obtained from sensors unit 120 and meteorological information obtained from meteorological data source 124 to determine plant stress classifications. Server database 134 stores relevant information that can be retrieved and processed by server processor 132.

[0046] Plant stress determination system 105 may further include at least some portions of user device 140. User device processor 144 performs data processing required by user device 140 and may receive instructions or data from other components of system 105. User device 140 may include a user management application 145 operating on user device processor 144. User management application 145 may provide information relating to determined plant stress received from server 130. In some cases, the user management application 145 may be implemented as a web application accessible via a web browser or as a mobile application configured to operate directly on a mobile device such as a smartphone. Information may be stored in a local memory (not shown) of user device 140. User interface 146 allows a user to receive information and to control settings of user device 140. User interface 146 may include a display screen configured to present visual content, such as alerts or reports issued by user management application 145 relating to plant stress determined by system 105. User interface 146 may be a graphical user interface, such as a cursor and / or a touch-screen menu interface. User interface 146 may also include peripheral communication devices configured to provide audible communication, such as a microphone and an audio speaker, as well as voice recognition capabilities to enable a user to enter instructions or data by means of speech commands.

[0047] Sensors unit 120, meteorological data source 124, server 130, and user device 140 may be communicatively coupled through communications network 128 to enable data transmission between system components. Information may be conveyed between elements of system 105 over any suitable data communication channel or network, using any type of channel or network model and any data transmission protocol.

[0048] The components and devices of plant stress determination system 105 may be based in hardware, software, or combinations thereof. It is appreciated that the functionality associated with each of the devices, components or modules of network environment 100 or system 105 may be distributed among multiple devices, components or modules, which may reside at a single location or at multiple locations. For example, the functionality associated with processors 132, 144 may be integrated, or may be distributed between multiple processing units. For example, the functionality of plant stress analysis module 133 may operate on one or more servers 130 and / or on user device 140. Network environment 100 and system 105 may optionally include and / or be associated with additional components or modules not shown in Figures 1, for enabling the implementation of the disclosed subject matter.

[0049] Reference is made to FIG. 2, which illustrates a schematic diagram of information flow in the plant stress determination system 105 of FIG. 1, according to aspects of the present disclosure. Plant stress analysis module 133 receives sap flow data 151 from sensors unit 120, where sap flow data 151 represents measurements obtained from sap flow sensors 122 monitoring respective plants at plant growth site 110. Plant stress analysis module 133 also receives ETO data 153 and meteorological data 155 from meteorological data source 124, where ETO data 153 reflects water loss dynamics of monitored plants, and meteorological data 155 includes atmospheric parameters such as temperature, humidity, wind speed, and precipitation.

[0050] Plant stress analysis module 133 includes an ETO threshold determination submodule 135 configured to process the meteorological data 155. ETO threshold determination submodule 135 generates an ETO threshold table 157 that defines ETO thresholds for different plants beyond which sap flow measurements may be characterized by irregular behavior. In some cases, ETO threshold determination submodule 135 trains decision tree models using meteorological data 155 and generates partial dependence plots to identify threshold values where sap flow behavior becomes irregular.

[0051] Plant stress analysis module 133 includes a false positive exclusion submodule 136 that receives the ETO threshold table 157 from ETO threshold determination submodule 135. False positive exclusion submodule 136 provides information for reducing false positive stress events based on conditions that exceed the defined ETO thresholds of ETO threshold table 157. In some cases, false positive exclusion submodule 136 identifies and filter out data collected during rainfall periods, from unstable sensors, or during periods where ETO exceeds predefined thresholds. Plant stress analysis module 133 includes a data preprocessing submodule 137 that receives the sap flow data 151 and the ETO data 153. Data preprocessing submodule 137 may apply preprocessing operations to the received data, such as filtering, smoothing, and normalization, to reduce noise and provide uniform properties for ensuring comparability of datasets across different conditions. Data preprocessing submodule 137 may perform additional preprocessing operations to facilitate subsequent processing by other submodules.

[0052] Plant stress analysis module 133 includes a stress classification submodule 138 that receives the processed data signals from data preprocessing submodule 137 along with false positive exclusion data from false positive exclusion submodule 136. Stress classification submodule 138 may determine plant stress data 159 based on a correlation analysis between preprocessed sap flow and ETO data. In some cases, stress classification submodule 138 may calculate correlation coefficients, generate stress scores, aggregate multiple stress scores into group scores, and apply dynamic thresholds to establish stress level classifications. It is noted that submodules 135, 136, 137, 138 are depicted for exemplary purposes as separate functional elements but may be implemented at least partially as a cohesive module.

[0053] As further shown in FIG. 2, user management application 145 may receive the plant stress data 159 from plant stress analysis module 133. User management application 145 may process plant stress data 159 to provide information to users through user interface 146. In some cases, user management application 145 issues an alert 147, such as when a severity or urgency of a stress level classification exceeds a predefined metric. User management application 145 may also generate a report 149 relating to plant stress data 159. Report 149 may include a list of stress level classifications for multiple plants at the plant growth site 110 in real-time or over a selected period.

[0054] Referring to FIG. 3, a method for plant stress determination may be implemented using the plant stress determination system 105. The method may be generally divided into three main phases: a data collection phase 210, an ETO threshold determination phase 220, and a plant stress determination phase 230. Each phase performs specific functions that contribute to the overall analysis and classification of plant stress conditions.

[0055] The data collection phase 210 involves gathering various types of data from the monitored plants and their environment. During data collection phase 210, the method collects the sap flow data 151 from sensors unit 120, where sap flow data 151 represents measurements obtained from sap flow sensors 122 monitoring respective plants at plant growth site 110. Data collection phase 210 also involves obtaining the ETO data 153 from meteorological data source 124, where ETO data 153 reflects evapotranspiration information or water loss dynamics of monitored plants. Additionally, data collection phase 210 may include collecting the meteorological data 155 from meteorological data source 124, where meteorological data 155 may encompass atmospheric parameters such as temperature, humidity, wind speed, and precipitation.

[0056] The ETO threshold determination phase 220 may follow data collection phase 210 and establishes threshold values for filtering irregular sap flow behavior. ETO threshold determination phase 220 includes a step 222 of decision tree training, where decision tree models may be trained to predict sap flow based on the meteorological data 155. ETO threshold determination phase 220 includes a step 224 for generating partial dependence plots, where partial dependence plots may be created for each decision tree model to analyze relationships between ETO and sap flow. ETO threshold determination phase 220 includes a step 226 for threshold identification, where the partial dependence plots are examined to identify ETO threshold values beyond which sap flow behavior becomes irregular. ETO threshold determination phase 220 may conclude with a step 228 for threshold table creation, where the identified ETO thresholds are systematically recorded in ETO threshold table 157 for different plants.

[0057] The plant stress determination phase 230 includes two main processes that operate together to classify plant stress conditions while minimizing false positive detections. Specifically, plant stress determination phase 230 includes a false positive elimination process 240 and a stress detection process 250. The false positive elimination process 240 identifies and excludes conditions that may lead to incorrect stress classifications, while the stress detection process 250 analyzes the filtered data to determine actual plant stress levels.

[0058] False positive elimination process 240 includes steps for filtering out data that may cause inaccurate stress determinations. False positive elimination process 240 includes a step 242 of rain filtering, where data collected during rainfall periods is identified and excluded from analysis. False positive elimination process 240 includes a step 244 of a sensor stability check, where data from unstable sap flow sensors 122 is filtered out based on predefined stability criteria. False positive elimination process 240 also includes a step 246 of a thresholds check, where data collected during periods when ETO exceeds the predefined ETO thresholds from ETO threshold table 157 is excluded from stress analysis.

[0059] Stress detection process 250 analyzes the filtered data to determine plant stress classifications through multiple sequential steps. Stress detection process 250 begins with a step 252 of preprocessing, where the collected sap flow data 151 and ETO data 153 may undergo filtering, smoothing, and normalization operations. Stress detection process 250 continues with a step 254 for correlation analysis, where correlation coefficients between the preprocessed sap flow data 151 and ETO data 153 are calculated. Stress detection process 250 includes a step 256 for daily stress score generation, where stress scores may be determined periodically based on the calculated correlation coefficients. Stress detection process 250 further includes a step 258 of block aggregation, where multiple individual stress scores are aggregated into group stress scores for blocks of plants with common properties. Stress detection process 250 includes a step 262 of dynamic thresholds processing, where dynamic thresholds are applied to the group stress scores to account for various external factors such as seasonal variations or historical data patterns. Stress detection process 250 concludes with a step 264 for stress level classification, where final stress level classifications are established based on the processed group stress scores, providing categorizations such as high stress, medium stress, low stress, or no stress conditions for the monitored plants.

[0060] Reference is now made to FIG. 4, which illustrates a detailed flow diagram of data collection phase 210 and ETO threshold determination phase 220 of the method of FIG. 3, according to aspects of the present disclosure. Data collection phase 210 and ETO threshold determination phase 220 may be implemented through a series of steps that gather and process information for plant stress analysis.

[0061] Data collection phase 210 begins with a step 212 for obtaining sap flow measurements from monitored plants at the plant growth site 110. During step 212, sap flow measurements of monitored plants at plant growth site 110, such as first plant 112, second plant 113, and third plant 114, are obtained from respective sap flow sensors 122 of sensors unit 120. Each sap flow sensor 122 may obtain measurements of sap flux density, which may be influenced by stomatai conductance of the monitored plants. The sap flow measurements may be obtained intermittently at periodic intervals and under varying environmental conditions, where each sap flow sensor 122 may be adapted to match characteristics or properties of the respective monitored plant.

[0062] Data collection phase 210 continues with a step 214 for obtaining evapotranspiration data. During step 214, ETO data 153 reflecting water loss dynamics of monitored plants at plant growth site 110 is obtained from meteorological data source 124. ETO data 153 may be obtained intermittently at periodic intervals, such as hourly, daily, weekly, and / or monthly. In some cases, ETO data 153 may be obtained from a publicly accessible weather / climate API. It is noted that ETO represents one example of meteorological data indicative of water dynamics of the monitored plants. In other examples, plant stress determination system 105 may collect alternative meteorological data indicative of plant water status, instead of or in addition to ETO data. In a further step 216 of data collection phase 210, additional meteorological data is obtained. During step 216, meteorological data 155 of plant growth site 110 is obtained from meteorological data sources 124. Meteorological data 155 may include atmospheric parameters associated with monitored plants 112, 113, 114 and / or plant growth site 110, such as: temperature, barometric pressure, humidity, wind speed, wind direction, vapor pressure deficit (VPD), precipitation, solar radiation, and the like. Such atmospheric parameters may provide further understanding of factors influencing sap flow of monitored plants 112, 113, 114. Meteorological data 155 may be obtained intermittently at periodic intervals, such as hourly, daily, weekly, and / or monthly.

[0063] ETO threshold determination phase 220 may establish ETO thresholds beyond which sap flow measurements behave irregularly for each respective plant. The ETO thresholds may indicate conditions that do not result from plant stress factors and may be used for filtering out time periods that should not be classified as plant stress events.

[0064] ETO threshold determination phase 220 begins with a step 222 of training a decision tree model for predicting sap flow based on received meteorological data. During step 222, ETO threshold determination submodule 135 receives meteorological data 155 and trains a decision tree learning model to predict sap flow for each respective plant. For example, a reference training dataset is formed based on a large quantity of reference plants, where each plant is labelled with measured sap flow output for a given set of meteorological data input. The training dataset, representing classification labels for multiple plants (or plant types), is iteratively trained using machine learning or supervised learning processes during a learning stage, to implicitly identify patterns and generate a decision tree model that can be applied on new input data for predictive classification during a subsequent analysis stage. The training process may apply data mining techniques including feature selection, cross-validation, and hyperparameter optimization to produce mapping functions that can be used for classifying additional instances of new datasets (meteorological data and sap flow measurements) of different plants (or plant types) according to relevant classification criteria. The decision tree model training may utilize alternative machine learning algorithms beyond decision tree approaches, including ensemble methods such as random forests or gradient boosting. In some cases, the training may utilize artificial neural network processes, such as convolutional neural networks, recurrent neural networks, or deep learning algorithms with multiple hidden layers for complex pattern recognition. The training may also employ classification or regression analysis approaches, such as linear regression models, logistic regression models, support-vector machine models, or kernel-based methods. Additionally, the training may implement decision tree learning approaches, such as random forest classifiers, or any combination of these machine learning techniques with appropriate regularization and validation procedures. The data analysis may utilize any suitable tool or platform, such as publicly available open-source machine learning or supervised learning tools.

[0065] A next step 224 of ETO threshold determination phase 220 includes generating partial dependence plots for each respective decision tree model using statistical analysis techniques. During step 224, ETO threshold determination submodule 135 generates a partial dependence plot (PDP) for each generated decision tree model, where the PDP may indicate the marginal effect of a selected input feature on the predicted output of the decision tree model through mathematical modeling. The PDP may represent the relationship between a target response, such as predicted sap flow, and a set of input features of interest, such as ETO, marginalized over other input features, such as additional meteorological data 155, using statistical integration methods. A PDP can be considered to represent an expected target response as a function of the input features of interest, computed by averaging the model predictions over the distribution of all other features. Accordingly, the generated partial dependence plots can provide an indication for how changes in ETO influence sap flow for identifying irregular patterns or correlations through visual and quantitative analysis. In other examples, the partial dependence plot may alternatively be embodied by a different model or data structure representing a relationship between an expected target response as a function of input features of interest, such as accumulated local effects plots or individual conditional expectation plots, in which irregularities may be identified.

[0066] ETO threshold determination phase 220 includes a further step 226 of identifying ETO thresholds beyond which sap flow behavior becomes irregular. During step 226, each generated PDP is examined to identify one or more points where the relationship between ETO (input feature of interest) and sap flow (target response) begins to exhibit significant changes. Such PDP points may represent thresholds beyond which the sap flow behavior may be considered irregular or substantially divergent from the relationship between the variables exhibited at other points of the partial dependence plot, such as determined by change point detection algorithms or slope analysis methods. For example, for a given PDP, the sap flow may be characterized below a certain ETO value by a consistent relationship or correlation, such as a substantially linear dependence with a stable slope, and characterized by an inconsistent relationship or correlation above the ETO value, which may be established as a threshold based on statistical criteria such as significant changes in derivative or variance. The PDP examination may be implemented manually, such as based on visual inspection by an expert technician using graphical tools, or automatically using suitable processing tools executed by ETO threshold determination submodule 135 including automated change point detection algorithms, breakpoint regression analysis, or machine learning-based anomaly detection methods.

[0067] ETO threshold determination phase 220 concludes with a step 228 for creating an ETO threshold table of identified ETO thresholds for each plant. During step 228, ETO threshold determination submodule 135 systematically records the identified ETO thresholds for different plants as a data table or equivalent data structure. The resultant ETO threshold table 157 may serve as a reference for filtering out non-stress time periods, to reduce false positives during the subsequent plant stress determination phase 230.

[0068] Reference is now made to FIG. 5, which illustrates a detailed flow diagram of plant stress determination phase 230 of the method of FIG. 3, according to aspects of the present disclosure. Plant stress determination phase 230 is implemented through the processing of collected data to classify plant stress conditions while minimizing false positive detections. Plant stress determination phase 230 includes a false positive elimination process 240 and a stress detection process 250. False positive elimination process 240 involves identifying and excluding conditions that may lead to incorrect stress classifications, while stress detection process 250 involves analyzing the filtered data to determine actual plant stress levels.

[0069] False positive elimination process 240 includes a step 242 of rain filtering, where data collected during rainfall periods is identified and filtered out. During step 242, false positive exclusion submodule 136 processes meteorological data 155 to detect the presence of rain based on predefined rain condition criteria. In some examples, rain conditions are detected if the amount or quantity of precipitation exceeds a minimum threshold over a selected time period. False positive exclusion submodule 136 may identify and flag sap flow data 151 collected during a rain condition period, where the flagged sap flow data 151 may be excluded from plant stress detection since rain may introduce noise into the sap flow signal. For example, the rain filtering filters out sap flow data 151 collected during a full day (e.g., a 24-hour period), or a shorter time period, such as one hour, when rain conditions are present. It is appreciated that the term “rain” used herein may broadly encompass different forms of precipitation, including but not limited to: drizzling, dew, snow, sleet, freezing rain, ice pellets, hail, graupel, and the like.

[0070] False positive elimination process 240 includes a step 244 of a sensor stability check, where collected data from unstable sap flow sensors is filtered out. During step 244, false positive exclusion submodule 136 identifies and flags sap flow data 151 obtained from sap flow sensors 122 that fail to pass a sensor stability check. The sensor stability check may be based on analysis of sensor output data parameters including number of measurements obtained during a selected duration, signal-to-noise ratio calculations, selected hours of measurements, signal amplitude, consistency, and temporal stability metrics. In some cases, rigorous stability checks may be performed regularly on sap flow sensors 122 using suitable analysis tools including statistical process control methods to ensure reliability of the collected sap flow data 151. The sensor stability check may assign an overall sensor stabilization score, such as a binary metric indicating whether the sensor is stable or unstable, or a continuous score reflecting the degree of stability, where false positive exclusion submodule 136 may direct stress classification submodule 138 to filter out flagged sap flow data 151 received from sensors that fail to meet predefined stability criteria. For example, submodule 136 may identify and filter out sap flow data collected from a sap flow sensor 122 that produces measurements below a predetermined signal-to-noise ratio (SNR) threshold over a selected period, or exhibits excessive drift or inconsistent readings that indicate sensor malfunction or degradation.

[0071] False positive elimination process 240 includes a step 246 of a thresholds check, where collected data during periods when ETO exceeds predefined ETO thresholds is filtered out. During step 246, false positive exclusion submodule 136 identifies and flags sap flow data 151 obtained during a common time period as ETO data 153 for the same plant that exceeds an ETO threshold for that plant specified in the ETO threshold table 157. False positive exclusion submodule 136 may process ETO threshold table 157 to detect if collected ETO data 153 for a monitored plant includes an ETO that exceeds a predefined ETO threshold for that plant beyond which the sap flow signal may exhibit irregular behavior. The ETO threshold filtering may filter out sap flow data 151 for a same day or shorter time periods such as specific hours when ETO exceeds the predefined thresholds, where false positive exclusion submodule 136 may direct stress classification submodule 138 to exclude the flagged sap flow data 151 from plant stress detection due to potential noise in the relevant sap flow signal.

[0072] As further shown in FIG. 5, stress detection process 250 analyzes the filtered data to determine plant stress classifications through multiple sequential steps that process the collected data while excluding false positive stress conditions. Stress detection process 250 begins with a step 252 of preprocessing, where collected data undergoes preprocessing operations to improve data quality and consistency. During step 252, data preprocessing submodule 137 receives and processes collected sap flow data 151, ETO data 153, and meteorological data 155 using digital signal processing methods. Data preprocessing submodule 137 may apply a first set of initial processing operations to the collected data, such as filtering using low-pass or band-pass filters and smoothing using moving averages or Gaussian filters, for reducing noise and minimizing signal artifacts characterizing short-term fluctuations rather than consistent trends. Submodule 137 may apply further processing operations, including normalization using z-score standardization or min-max scaling, to provide uniform properties of the collected data for ensuring comparability of the datasets across different conditions and measurement scales. At least some of the preprocessing operations may be applied at regular intervals, such as a daily basis, and may include outlier detection and removal using statistical methods to ensure data quality.

[0073] Stress detection process 250 continues with a step 254 of correlation analysis, where correlation coefficients between the preprocessed sap flow data and ETO data is determined. During step 254, stress classification submodule 138 receives the preprocessed datasets and determines at least one correlation coefficient between the sap flow data 151 and the ETO data 153 to quantify the strength and direction of their relationship. The correlation coefficients may include one or more of a Pearson correlation coefficient (e.g., for measuring linear relationships), a Spearman rank correlation coefficient (e.g., for capturing monotonic relationships), and a Kendall rank (Kendall's tau) correlation coefficient (e.g., for assessing concordance between variables). The correlation coefficients may be selected to provide a robust measure of the relationship between the two variables, such as by capturing linear and non-linear associations between the sap flow data 151 and the ETO data 153. In some cases, multiple correlation measures may be computed and combined using weighted averaging or ensemble methods to provide a more comprehensive assessment of the relationship. The correlation analysis may also include time- lagged correlations to account for potential delays between evapotranspiration changes and sap flow responses, and may incorporate confidence intervals and statistical significance testing to ensure reliability of the correlation measurements.

[0074] Stress detection process 250 includes a further step 256 of daily stress score generation, where stress scores are determined periodically based on the correlation coefficients through quantitative analysis. During step 256, stress classification submodule 138 determines a stress metric or score for the monitored plant based on the correlation coefficients through quantitative analysis. The plant stress score may be determined based on a mathematical function of the correlation coefficients such as calculating an average, weighted averaging, or more complex functions that account for the relative importance of different correlation measures. The determined stress score represents an overall stress level for the monitored plant expressed as a quantitative measure. For example, the stress score may be a numerical quantity between 0 and 1, where a lower value corresponds to a higher degree or severity of plant stress indicating poor correlation between sap flow and evapotranspiration, and a higher value corresponds to a lower degree or severity of plant stress indicating strong correlation. The stress scores may be determined for selected time periods such as daily or hourly intervals, and may incorporate temporal smoothing or filtering to reduce noise and provide stable measurements over time.

[0075] Stress detection process 250 may include a next step 258 of block aggregation, where a plurality of determined stress scores is aggregated into a respective group. During step 258, stress classification submodule 138 determines a group stress score for a group of plants based on an aggregation of multiple individual stress scores for individual plants using mathematical functions that account for data quality and reliability. A group stress score may be determined for a designated set of plants with common properties (also referred to herein as a “block”), such as a group of plants that are handled collectively at plant growth site 110, such as an irrigation block of multiple plants that are irrigated collectively. For example, an irrigation block may be composed of a plurality of plants 112, 113, 114 in a defined region of plant growth site 110, where a respective sap flow sensor 122 is coupled to each plant in the block. A group (e.g., block level) stress score may be determined based on a mathematical function of multiple individual (e.g., unit level) stress scores with reliability metrics affecting proportional weighting in the calculation through weighted averaging algorithms. Each individual stress score may be associated with a reliability metric reflective of a reliability or confidence level of the respective score determination, such as based on the reliability of the underlying data including sensor performance metrics, data completeness, and signal quality indicators, where individual stress scores having a higher reliability metric may be assigned a higher proportional weighting in a group stress score calculation. For example, a first stress score having a high confidence level (e.g., based on high reliability sap flow measurements with good signal-to-noise ratio) is assigned a first weighting which is greater than a second stress score having a low confidence level (e.g., based on less reliable sap flow measurements with poor signal quality). A group stress score may represent a correlation between a periodic ETO to periodic sap flow measurements for a group of plants, such as a numerical quantity between 0 and 1, where a lower value corresponds to a higher severity of plant stress. The group stress score may be determined at selected times and intervals, such as daily or hourly, and may include statistical measures such as confidence intervals to indicate the reliability of the aggregated score.

[0076] Stress detection process 250 includes a next step 262 of dynamic thresholds processing, where dynamic thresholds are applied to respective group stress scores. During step 262, stress classification submodule 138 modifies a group stress score using one or more dynamic thresholds to account for various factors that may influence stress determination accuracy. Dynamic stress score thresholds may be established for each plant or plant type using statistical analysis of historical data, where each dynamic stress score threshold is associated with an external factor or condition that may influence stress determination. An updated group stress score modified according to these dynamic thresholds may thereby be reflective of such underlying factors or conditions through mathematical adjustments. In some examples, different dynamic thresholds are established based on factors such as time of year (e.g., different thresholds for different calendrical months based on seasonal plant behavior), seasonal variations (e.g., different thresholds for different seasons accounting for natural physiological changes), or historical data patterns (e.g., based on previous stress scores for the respective plant or plant type using statistical baselines). The dynamic thresholds may be established using adaptive thresholding processing techniques, such as utilizing historical stress score data as reference through statistical modeling, and may be repeatedly updated according to new data and changing conditions using machine learning algorithms that adapt to evolving environmental patterns.

[0077] Stress detection process 250 concludes with a step 264 of stress level classification, where stress level classifications are established according to the group stress scores. During step 264, stress classification submodule 138 determines a stress level classification for the monitored plant based on the group stress score, optionally modified according to dynamic thresholds accounting for external conditions, through comparison with established threshold values. The determined stress level classification may represent a degree or severity of plant stress experienced by the monitored plant, such as a numerical quantity between 0 and 1 where a lower value corresponds to a higher degree of severity of plant stress indicating poor physiological performance. In some examples, the stress level classification includes categories of high stress, medium stress, low stress, or no stress, with corresponding numerical ranges that may be calibrated based on empirical data and expert knowledge. An exemplary plant stress classification may be as follows: a high stress or severe state, corresponding to a stress score less than 0.4 indicating significant physiological impairment; a medium stress or mild state, corresponding to a stress score of at least 0.4 and less than 0.6 indicating moderate physiological stress; a low stress or normal state, corresponding to a stress score of at least 0.6 and less than 0.8 indicating minor physiological stress; and a no stress or ideal state, corresponding to a stress score of at least 0.8 and less than 1.0 indicating optimal physiological performance. These threshold values may be adjusted based on plant type, environmental conditions, and historical performance data to optimize classification accuracy. The stress level classification may be established periodically according to a periodic group level stress score, such as once a day based on a daily determined group stress score, and the classification may be repeatedly updated according to new data and changing conditions using real-time processing algorithms that ensure timely detection of stress events.

[0078] Referring back to FIGS. 1 and 2, user management application 145 operating on user device 140 may receive plant stress data 159 from plant stress analysis module 133. Plant stress data 159 may include periodically determined stress levels for a plurality of plants 112, 113, 114 at plant growth site 110. Application 145 may be configured to receive a restricted subset of plant stress data 159 based on predefined criteria, such as access privileges associated with user device 140. For example, selected users may be granted access to specific information, such as an authorized operator (e.g., a supervisor) of plant growth site 110.

[0079] Application 145 may issue a notification or alert 147 relating to plant stress data 159. For example, an alert 147 may be issued when a severity and / or urgency of a stress level classification exceeds a predefined metric. An alert 147 may be issued when a stress level of a plant is classified as a “high stress or severe state”, or when the plant is determined to require urgent attention to avoid irreparable harm from plant stress factors. The alert 147 may be displayed on user interface 146 through visual indications (e.g., text messages, displayed markings, symbols or other graphical information) and / or audible indications (e.g., alarms, beeps, buzzers, ringtones). Instructions for issuing an alert 147 may be sent from server 130 to a plurality of applications 145 operating on respective devices 140 of multiple users, and / or to other destinations, such as a service center or monitoring station associated with plant growth site 110. An alert 147 relating to plant stress data 159 may be provided in the absence of a dedicated application 145, such as through a text messaging service or instant messaging platform received by user device 140.

[0080] Application 145 may generate a report 149 relating to plant stress data 159. Report 149 may include a list of stress level classifications for multiple plants at plant growth site 110 in realtime and / or over a selected period. Different plants may be grouped into categories or sub-categories based on plant type (taxonomy) and / or stress level classification. For example, plants of a first plant type having a high stress level classification for at least a minimum period (e.g., over three consecutive days) may be grouped into a first category designated for high priority (urgent) attention. Plants of the first plant type having a medium stress level classification for at least a minimum period may be grouped into a second category designated for low priority attention. Report 149 may further include statistics associated with determined stress levels, such as historical data obtained for the same or similar plants, plant types, and / or plant growth sites at previous dates and times. The information and statistics may be reported over a selected duration, depicting changes in determined stress levels over days, weeks, months, years, or other time periods, such as calendrical or agricultural seasons.

[0081] Following determination of plant stress data 159 (and possible alert 147 or report 149 relating thereto), additional measures may be deployed to identify plant stress factors, such as identifying one or more factors potentially causing stress in a monitored plant 112, 113, 114 classified as having a high stress or severe state. For example, a designated user may be dispatched to manually examine a monitored plant to determine potential stress factors (e.g., water deficit, nutrient deficiency, temperature extreme, high salinity, disease, pathogen). Stress factor identification may be performed using suitable tools and processes, such as image processing or experimental analysis. A temperature sensor may be utilized to verify a temperature-related stress condition factor, and laboratory testing may be utilized to detect a plant disease or pathogen.

[0082] A user may perform at least one stress mitigation action in response to the determined plant stress data 159. In particular, a user may implement suitable measures to address, minimize, or overcome stress factors, and / or treat stress symptoms of at least one plant, such as a monitored plant classified as having a high stress or severe state, thereby improving cultivation at plant growth site 110. A user may utilize the provided report 149 of plant stress data 159 to decide on responsive measures, such as determining the timing and nature of stress mitigation actions to be implemented (e.g., which actions should be taken, which plants should be prioritized). Application 145 may also provide recommendations or suggested actions for stress mitigation in accordance with the determined plant stress data 159. Such recommendations may be based on user feedback. For example, a user may specify certain constraints for applying measures at plant growth site 110, or designate selected plants or plant types as higher priority for stress mitigation, and application 145 may determine and present recommendations in accordance with the user-specified criteria.

[0083] Plant growth site 110 may undergo plant stress monitoring during selected time intervals and / or on a continuous basis. Results of stress mitigation measures and determined plant stress data may be evaluated at subsequent times to improve further stress mitigation. For example, stress mitigation actions may be applied at plant growth site 110 in response to a first set of plant stress data obtained at a first time, and a second set of plant stress data obtained at a second time may be examined to evaluate the progress or success rate of the applied mitigation actions, and to suggest further or alternative actions if necessary. Plant stress data, including dynamic behavior and statistics, obtained over subsequent time periods may provide a more comprehensive evaluation to allow for targeted recommendations for enhancing plant cultivation at plant growth site 110. Determined plant stress data 159 at a first plant growth site 110 may also be applied to enhance plant cultivation at other plant growth sites, such as at a second growth site having similar plants, plant types and / or other properties as the first growth site. For example, successful stress mitigation measures applied at the first growth site may be implemented at the second growth site.

[0084] The present disclosure may provide improved stress determination of plants reflective of various stress factors not limited to water-related issues. The disclosed method and system may be deployed at various plant growth sites containing a variety of plant types. The disclosed plant stress determination system may require limited physical components and may thus be deployed with minimal installation costs and time. The system elements may not be reliant on frequent monitoring and maintenance and do not require constant calibration. The disclosed method and system may not require specialized equipment or uniquely skilled personnel. The determined plant stress data may be dynamically and adaptively updated in accordance with changing real-world conditions. Alerts and reporting may be provided to relevant users to allow for implementation of suitable measures to mitigate detected plant stress and ensure effective plant cultivation.

[0085] In some embodiments, the plant stress determination system 105 may be implemented as computer-executable instructions stored on a non-transitory computer-readable storage medium. The computer-readable storage medium may include any suitable storage device capable of storing instructions that, when executed by a processor, cause the processor to perform the plant stress determination operations described herein. Examples of suitable computer-readable storage media may include, but are not limited to, magnetic storage devices such as hard disk drives, optical storage devices such as compact discs (CDs) or digital versatile discs (DVDs), solid-state storage devices such as flash memory or solid-state drives (SSDs), and other forms of non-volatile memory. The instructions may be embodied as software applications, firmware, or other executable code that implements the functionality of the plant stress analysis module 133, including the various submodules and processing steps described throughout this disclosure.

[0086] The computer-readable storage medium may contain instructions that, when executed, cause a processor to perform operations including obtaining sap flow measurements from at least one plant, receiving evapotranspiration data from meteorological data sources, analyzing correlations between the sap flow measurements and evapotranspiration data, generating stress indicators based on the analyzed correlations, and determining stress conditions of monitored plants. The instructions may further cause the processor to perform preprocessing operations such as filtering, smoothing, and normalization of collected data, as well as implementing false positive elimination processes including rain filtering, sensor stability checks, and threshold-based filtering. Additionally, the instructions may enable the processor to perform aggregation of multiple stress indicators into group stress scores, apply dynamic thresholds based on external factors, and generate stress level classifications for monitored plants.

[0087] In some aspects, the plant stress determination system 105 may be embodied as a plant monitoring apparatus specifically configured for field deployment at plant growth sites. The plant monitoring apparatus may include a housing designed to withstand outdoor environmental conditions and protect internal components from moisture, dust, and temperature variations. The housing may be constructed from weather-resistant materials such as reinforced plastics, metals with protective coatings, or composite materials suitable for agricultural environments. The housing may include mounting features or attachment mechanisms that allow the apparatus to be securely positioned at the plant growth site, such as stakes for ground mounting, clamps for attachment to support structures, or brackets for wall or pole mounting.

[0088] The plant monitoring apparatus may include at least one sap flow sensor 122 mounted within or integrated into the housing. The sap flow sensor 122 may be configured to be positioned in direct contact with plant tissue, such as tree bark or stem surfaces, to obtain accurate sap flow measurements. The sensor mounting may include adjustable positioning mechanisms to accommodate different plant sizes and geometries. The apparatus may include a communication interface configured to establish data communication with external systems, such as meteorological data sources 124, server 130, or user devices 140. The communication interface may support various communication protocols and technologies, including wireless communication such as Wi-Fi, cellular networks, or satellite communication, as well as wired communication options such as Ethernet or serial connections.

[0089] The plant monitoring apparatus may further include a processing unit operatively coupled to the sap flow sensor 122 and the communication interface. The processing unit may be embodied as a microprocessor, microcontroller, digital signal processor, or other suitable computing device capable of executing instructions and performing data processing operations. The processing unit may be configured to analyze correlations between sap flow measurements and evapotranspiration data, generate stress indicators based on the analyzed correlations, and determine stress conditions of monitored plants. The processing unit may implement the functionality of the plant stress analysis module 133 and its various submodules, including data preprocessing, false positive elimination, and stress classification operations. The apparatus may also include an output interface configured to provide information relating to determined stress conditions. The output interface may include visual indicators such as LED displays, LCD screens, or status lights, audible indicators such as speakers or buzzers, and data communication capabilities for transmitting stress information to remote systems or user devices.

[0090] The plant monitoring apparatus may be designed for autonomous operation in field conditions, requiring minimal maintenance and user intervention. The apparatus may include power management systems such as rechargeable batteries, solar panels, or external power connections to ensure continuous operation. The apparatus may also include environmental protection features such as sealed enclosures, drainage systems, and temperature regulation to maintain reliable operation across varying weather conditions. In some cases, multiple plant monitoring apparatus units may be deployed across a plant growth site 110 to monitor different plants or different areas, with each apparatus capable of independent operation while also supporting coordinated monitoring through network communication with other apparatus units and central monitoring systems.

[0091] Referring to Figures 6A-6B, the accuracy of plant stress detection using the plant stress determination system 105 of the present disclosure may be compared to plant stress detection using a pressure chamber technique for different grape variants. The comparison may demonstrate the effectiveness of the disclosed sap flow-based approach for determining plant stress conditions in agricultural applications. The pressure chamber technique may serve as a reference measurement method, where the pressure chamber may measure water potential of plant tissues by placing a detached leaf or shoot in a sealed chamber and introducing compressed gas until sap begins to exude from the cut end.

[0092] As shown in FIG. 6 A, a first graph 310 depicts plant stress measurements for a first grape variant over time. The first graph 310 includes a plot 312 representing plant stress measurements obtained using the pressure chamber technique, and a plot 314 representing plant stress measurements obtained using aspects of the disclosed system and method. The first graph 310 includes vertical lines indicating irrigation events, where a first irrigation time 316 and a second irrigation time 317 mark specific times when irrigation was applied to the monitored plants. The trends of plot 312 and plot 314 generally correspond to one another, with both plots exhibiting a gradual increase in stress levels, followed by a sharp decrease prior to the first irrigation time 316, and then a subsequent sharp increase after irrigation. With reference to FIG. 6B, a second graph 320 depicts plant stress measurements for a second grape variant over time. The second graph 320 includes a plot 322 representing pressure chamber plant stress measurements, and a plot 324 representing plant stress measurements obtained using aspects of the disclosed system and method. The second graph 320 includes vertical lines indicating irrigation events, where a first irrigation time 326 and a second irrigation time 327 mark specific irrigation application times. The trends of plot 322 and plot 324 generally correspond to one another, with both plots remaining fairly stable initially, followed by a small increase prior to the first irrigation time 326, and then a subsequent sharp decrease after irrigation application. The comparison shown in graphs 310 and 320 may demonstrate that the plant stress determination system 105 of the present disclosure may provide plant stress measurements that correlate well with established pressure chamber techniques across different grape variants. The disclosed system and method may offer advantages over pressure chamber measurements, such as continuous monitoring capabilities without requiring destructive sampling of plant tissues. The sap flow-based approach of the disclosed system and method may provide real-time stress detection without the need for trained personnel to visit the field and implement manual measurements, while the pressure chamber technique may require cutting plant portions and specialized expertise for accurate operation, and is limited to the number of plants that can be examined.

[0093] Referring to FIG. 7, a graph 340 illustrates the relationship between plant transpiration, rainfall, and stress levels over time for an almond plant. Graph 340 includes a transpiration plot 342 that represents actual consumption measurements overtime, with the measurements expressed in millimeters. Transpiration plot 342 shows the water consumption patterns of the monitored almond plant, which varies according to environmental conditions and plant physiological responses. Transpiration plot 342 demonstrates how water consumption changes in response to various factors such as temperature, humidity, and plant stress conditions. Graph 340 further includes rainfall indicators 344 that indicate precipitation amounts over the same time period as the transpiration measurements. Rainfall indicators 344 are represented as vertical bars with heights corresponding to the amount of precipitation measured in millimeters. Rainfall indicators 344 show the timing and intensity of rainfall events that may affect plant water availability and stress conditions. In some cases, rainfall indicators 344 may correspond to the rain filtering performed in step 242 of false positive elimination process 240, where false positive exclusion submodule 136 may identify rainfall periods and filter out corresponding sap flow data 151 to prevent false positive stress detections during precipitation events. Graph 340 further includes a stress indicator bar 346 that extends horizontally across the graph 340 to show plant stress levels over time. Stress indicator bar 346 displays different shadings or colorations that represent varying levels of plant stress, such as low stress, medium stress, and high stress conditions. The different shadings of the stress indicator bar 346 may correspond to the stress level classifications determined by stress classification submodule 138 during step 264 of stress detection process 250. Stress indicator bar 346 provides a visual representation of how plant stress varies in relation to the transpiration plot 342 and the rainfall bars 344, demonstrating the correlations between water consumption, precipitation, and stress conditions.

[0094] Graph 340 demonstrates how plant stress levels change in response to transpiration patterns and irrigation events. The relationship shown in graph 340 indicates that stress levels may increase when transpiration decreases and when irrigation or rainfall ceases. Graph 340 further shows that periods of low transpiration may correspond to higher stress levels as indicated by the stress indicator bar 346, while periods following rainfall events shown by the rainfall bars 344 may correspond to reduced stress levels. The correlations displayed in the graph 340 may validate the effectiveness of the plant stress determination system 105 in accurately detecting and classifying plant stress conditions based on the analysis of sap flow data 151 in relation to ETO data 153 and meteorological data 155 according to aspects of the present disclosure.

[0095] It will be appreciated that the embodiments described above are cited by way of example, and that the present disclosure is not limited to what has been particularly shown and described hereinabove. Rather, the scope of the present disclosure includes both combinations and sub-combinations of the various features described hereinabove, as well as variations and modifications thereof which would occur to persons skilled in the art upon reading the foregoing description and which are not disclosed in the prior art.

Claims

CLAIMS1. A system for determining plant stress, comprising: at least one sap flow sensor, configured to obtain sap flow measurements from at least one plant; a meteorological data source, configured to provide evapotranspiration data relating to the at least one plant; and a processor, configured to: receive the sap flow measurements and the evapotranspiration data, analyze correlations between the sap flow measurements and the evapotranspiration data, generate a stress indicator based on the analyzed correlations, and determine a stress condition of the at least one plant based on the stress indicator.

2. The system of claim 1, wherein analyzing correlations comprises determining at least one correlation coefficient between the sap flow measurements and the evapotranspiration data.

3. The system of claim 1 , wherein the processor is further configured to receive meteorological data and filter out sap flow measurements obtained during precipitation periods based on the meteorological data.

4. The system of claim 1, wherein the processor is further configured to evaluate sensor stability of the at least one sap flow sensor and filter out sap flow measurements obtained from sap flow sensors that fail to meet predefined stability criteria.

5. The system of claim 1, wherein the processor is further configured to: receive meteorological data; train at least one predictive model to predict sap flow based on the meteorological data; generate a dependence analysis for the predictive model to identify relationships between evapotranspiration and sap flow; identify an evapotranspiration threshold beyond which sap flow behavior becomes irregular based on the dependence analysis; and filter out sap flow measurements obtained during periods when the evapotranspiration data exceeds the identified evapotranspiration threshold.

6. The system of claim 5, wherein the dependence analysis comprises a partial dependence plot that indicates a marginal effect of evapotranspiration data on predicted sap flow.

7. The system of claim 1, wherein the processor is further configured to preprocess the sap flow measurements and the evapotranspiration data by applying at least one operation selected from the group consisting of: filtering; smoothing; normalization; and noise reduction.

8. The system of claim 1, wherein the at least one plant comprises a plurality of plants and wherein the processor is further configured to: generate a plurality of stress indicators corresponding to the plurality of plants; aggregate the plurality of stress indicators into a group stress indicator; and apply at least one dynamic threshold to the group stress indicator based on external factors.

9. The system of claim 8, wherein aggregating the plurality of stress indicators comprises: determining reliability metrics for individual stress indicators based on reliability of underlying sap flow measurements; and applying proportional weighting based on the reliability metrics when calculating the group stress indicator.

10. The system of claim 8, wherein the external factors comprise at least one of: time of year; seasonal variations; historical data patterns; and plant type characteristics.

11. The system of claim 1, wherein the at least one sap flow sensor is configured to determine sap flow based on thermal heat transfer analysis using a heat source and a thermal sensor positioned in proximity to plant tissue.

12. The system of claim 1, further comprising: a user device having a user interface; and a user management application configured to provide information relating to the determined stress condition through the user interface.

13. The system of claim 12, wherein the user management application is configured to provide at least one selected from the group consisting of: an alert when the determined stress conditions exceeds a predefined severity threshold; a report comprising stress level classifications for multiple plants over a selected period of time; historical trend analysis of stress conditions; and stress mitigation recommendations based on the determined stress condition.

14. The system of claim 1, wherein the stress condition comprises a classification selected from: high stress, medium stress, low stress, and no stress.

15. The system of claim 1, wherein the processor is further configured to: determine the stress indicator at periodic intervals; and update the stress condition determination based on changing environmental conditions and new sap flow measurements.

16. A method for determining plant stress, comprising: obtaining sap flow measurements from at least one plant using at least one sap flow sensor; receiving evapotranspiration data relating to the at least one plant from a meteorological data source; analyzing correlations between the sap flow measurements and the evapotranspiration data; generating a stress indicator based on the analyzed correlations; and determining a stress condition of the at least one plant based on the stress indicator.

17. The method of claim 16, wherein analyzing correlations comprises determining at least one correlation coefficient between the sap flow measurements and the evapotranspiration data.

18. The method of claim 16, further comprising receiving meteorological data and filtering out sap flow measurements obtained during precipitation periods based on the meteorological data.

19. The method of claim 16, further comprising evaluating sensor stability of the at least one sap flow sensor and filtering out sap flow measurements obtained from sap flow sensors that fail to meet predefined stability criteria.

20. The method of method of claim 16, further comprising: receiving meteorological data; training at least one predictive model to predict sap flow based on the meteorological data; generating a dependence analysis for the at least one predictive model to identify relationships between evapotranspiration and sap flow; identifying an evapotranspiration threshold beyond which sap flow behavior becomes irregular based on the dependence analysis; and filtering out sap flow measurements obtained during periods when evapotranspiration data exceeds the identified evapotranspiration threshold.

21. The method of claim 20, wherein the dependence analysis comprises a partial dependence plot that indicates a marginal effect of evapotranspiration data on predicted sap flow.

22. The method of claim 16, further comprising preprocessing the sap flow measurements and the evapotranspiration data by applying at least one operation selected from the group consisting of: filtering, smoothing, normalization, and noise reduction.

23. The method of method of claim 16, wherein the at least one plant comprises a plurality of plants, and wherein the method further comprises: generating a plurality of stress indicators corresponding to the plurality of plants; aggregating the plurality of stress indicators into a group stress indicator; and applying at least one dynamic threshold to the group stress indicator based on external factors.

24. The method of claim 23, wherein aggregating the plurality of stress indicators comprises: determining reliability metrics for individual stress indicators based on reliability of underlying sap flow measurements; andapplying proportional weighting based on the reliability metrics when calculating the group stress indicator.

25. The method of claim 23, wherein the external factors comprise at least one selected from the group consisting of: time of year, seasonal variations, historical data patterns, and plant type characteristics.

26. The method of claim 16, wherein obtaining sap flow measurements comprises determining sap flow based on thermal heat transfer analysis using a heat source and a thermal sensor positioned in proximity to plant tissue.

27. The method of claim 16, further comprising providing information relating to the determined stress condition through a user interface of a user device.

28. The method of claim 27, wherein providing information comprises providing at least one selected from the group consisting of: an alert when the determined stress conditions exceeds a predefined severity threshold; a report comprising stress level classifications for multiple plants over a selected period of time; historical trend analysis of stress conditions; and stress mitigation recommendations based on the determined stress condition.

29. The method of claim 16, wherein the stress condition comprises a classification selected from the group consisting of: high stress, medium stress, low stress, and no stress.

30. The method of claim 16, further comprising: determining the stress indicator at periodic intervals; and updating the stress condition determination based on changing environmental conditions and new sap flow measurements.

31. A non-transitory computer-readable storage medium containing instructions that, when executed by a processor, cause the processor to perform operations comprising:obtaining sap flow measurements from at least one plant using at least one sap flow sensor; receiving evapotranspiration data relating to the at least one plant from a meteorological data source; analyzing correlations between the sap flow measurements and the evapotranspiration data; generating a stress indicator based on the analyzed correlations; and determining a stress condition of the at least one plant based on the stress indicator.

32. A plant monitoring apparatus for determining plant stress, comprising: a housing configured for deployment at a plant growth site; at least one sap flow sensor mounted within the housing and configured to be positioned in contact with plant tissue to obtain sap flow measurements from at least one plant; a communication interface configured to receive evapotranspiration data from a meteorological data source; and a processing unit operatively coupled to the at least one sap flow sensor and the communication interface, the processing unit configured to: analyze correlations between the sap flow measurements and the evapotranspiration data, generate a stress indicator based on the analyzed correlations; determine a stress condition of the at least one plant based on the stress indicator; and an output interface configured to provide information relating to the determined stress condition.

Images