Modeling and prediction of below-ground performance of agricultural biological products in precision agriculture

Abstract

A below-ground agricultural biological performance modeling approach in precision agriculture combines customized field modeling with machine learning techniques for environmental matching of variables to describe a below-surface soil state, to understand and predict the performance of soil-active agricultural biological products such as bio-pesticides, bio-stimulants, plant growth regulators, and other biologically-derives soil adjuvants. The modeling approach characterizes the influence of environmental relationships on the performance of such soil-active agricultural biological products to develop a suite of predictive models to provide notifications, advisories, and recommendations for appropriate products for individual fields.

1 . A method, comprising:
 ingesting, as input data, above-ground weather data, represented by weather and climatological information that includes at least one of recent and current field-level weather data, historical field-level weather data and extended-range weather forecast data for a specific geographical area, crop-specific information relative to a crop in a particular field in the specific geographical area, soil-specific information relative to a soil system of the particular field, and a measured efficacy of a soil-active agricultural biological product in an outcome of one or more prior application of the soil-active agricultural biological product, wherein the soil-active agricultural biological product includes at least one of a bio-pesticide, a bio-stimulant, and a plant growth regulator;   modeling the input data in a plurality of data processing modules within a computing environment in which the plurality of data processing modules are executed in conjunction with at least one specifically-configured processor, the data processing modules configured to analyze a below-ground performance of the soil-active agricultural biological product, by   developing a customized field model of one or more physical and empirical characteristics impacting the below-ground performance of the soil-active agricultural biological product in the soil system, to enable a targeted application schedule at a desired time and a desired place to obtain a specific outcome, by   applying the input data to characterize the influence of environmental factors on the one or more prior applications of the soil-active agricultural biological product wherein the customized field model is configured to assess below-ground spatial and temporal soil system attributes that at least include soil moisture and temperature at multiple depths and over multiple time periods, with a specific soil-active agricultural biological product applied to the soil system in the particular field, and   correlating a significance of the above-ground weather data on the one or more prior applications of the soil-active agricultural biological product with the assessed below-ground spatial and temporal soil system attributes in the particular field to develop an environmental profile for the performance of the soil-active agricultural biological product at a below-surface state represented in the below-ground spatial and temporal soil system attributes, by selecting one or more appropriate environmental predictors impacting biological processes in the particular field, by a) constructing variables describing the below-surface state of the soil system in the particular field to summarize the correlated significance of the above-ground weather data and the assessed below-ground spatial and temporal soil system attributes, and b) matching the variables describing the below-surface state of the soil system in the particular field being modeled with the outcome in one or more prior applications of the soil-active agricultural biological product;   generating, as output data, a prediction of the below-ground performance of the soil-active agricultural biological product from the environmental profile; and   developing the targeted application schedule from the predicted below-ground performance for application of a specific soil-active agricultural biological product at desired time or at a desired geographical location, wherein a user applies the specific soil-active agricultural biological product based on the targeted application schedule at the desired time or at the desired geographical location, or an automated application of the specific soil-active agricultural biological product at the desired time or at the desired geographical location is controlled based on the targeted application schedule.   
     
     
         2 . The method of  claim 1 , further comprising applying the environmental profile to provide one or more advisories regarding an application of the soil-active agricultural biological product in any field at any selected time based on the prediction of the performance of the soil-active agricultural biological product. 
     
     
         3 . The method of  claim 1 , wherein the input data further includes field-specific data that includes positional information identifying the particular field, the positional information including GPS data that enables usage of GPS receivers to determine one or more of positional coordinates and boundaries of the particular field. 
     
     
         4 . The method of  claim 1 , wherein the input data further includes soil-water information for the particular field, the soil-water information including one or more of surface and sub-surface drainage characteristics, an amount of water applied to the soil system through effective rainfall and irrigation, an amount of water that is retained in a dynamically-growing root zone, an amount of water depleted from the dynamically-growing root zone, and irrigation-specific information relative to past, recent, or planned irrigation activity. 
     
     
         5 . The method of  claim 1 , where the soil-specific information includes one or more of a soil type, and known soil temperature and moisture content for at least one of a current growing season and a period of time extending up to and including a current growing season. 
     
     
         6 . The method of  claim 1 , wherein the crop-specific information further includes one or more of crop type data, planting data, crop growth data, growth stage-dependent parameters indicative of one or more crop growth stages, crop relative maturity data, crop planting depth and row spacing data, and targeted crop moisture or temperature thresholds. 
     
     
         7 . The method of  claim 1 , wherein the input data further includes field-level remotely-sensed imagery data of the particular field. 
     
     
         8 . The method of  claim 1 , wherein the correlating a significance of the above-ground weather data on the outcome of the one or more prior applications of the soil-active agricultural biological product with the assessed below-ground spatial and temporal soil system attributes in the particular field further comprises iteratively selecting the one or more appropriate environmental predictors to develop one or more geographically-dispersed machine learning base models for a particular soil-active agricultural biological product, and applying the one or more geographically-dispersed machine learning base models to environmentally match the particular soil-active agricultural biological product for a particular geographic area. 
     
     
         9 . The method of  claim 8 , further comprising training the one or more geographically-dispersed machine learning base models with performance data representing efficacy of the particular soil-active agricultural biological product for a smaller local region to re-learn the one or more geographically-dispersed machine learning base models for a specific localized geographical area within the smaller local region. 
     
     
         10 . The method of  claim 1 , further comprising automatically developing an artificial intelligence model to analyze time-varying characteristics of the soil system, by building a comprehensive dataset for the customized field model of one or more physical and empirical characteristics impacting the soil system, and applying the artificial intelligence model to predict a below-ground performance of a soil-active agricultural biological product in any field at any selected time. 
     
     
         11 . A method of evaluating a below-ground performance of a soil-active agricultural biological product in precision agriculture, comprising:
 developing a customized field model of one or more physical and empirical characteristics impacting below-ground performance of a soil-active agricultural biological product in a soil system in a particular field to enable a targeted application schedule at a desired time and a desired place to obtain a specific outcome, the customized field model configured to assess below-ground spatial and temporal soil system attributes that at least include soil moisture and soil temperature by at multiple depths and over multiple time periods, and with a specific soil-active agricultural biological product applied to the soil system in the particular field and characterize environmental factors influencing performance of a soil-active agricultural biological product in the soil system of the particular field, based on crop-specific information relative to a crop in the particular field, soil-specific information relative to the soil system of the particular field, and a measured efficacy in an outcome of one or more prior applications of the soil-active agricultural biological product, wherein the soil-active agricultural biological product includes at least one of a bio-pesticide, a bio-stimulant, and a plant growth regulator;   identifying one or more appropriate environmental predictors impacting biological processes in the particular field to develop an environmental profile for the performance of the soil-active agricultural biological product at a below-surface state represented in the below-ground spatial and temporal soil system attributes, and to analyze the below-surface state of the soil system in one or more machine learning models configured to construct variables describing the below-surface state and summarizing a correlated significance between above-ground weather data represented by weather and climatological information that includes at least one of recent and current field-level weather data, historical field-level weather data and extended-range weather forecast data for the specific geographical area and the below-ground spatial and temporal soil system attributes, and to match the variables summarizing a correlated significance with the one or more prior applications of the soil-active agricultural biological product;   predicting a performance of soil-active agricultural biological product based on the environmental profile; and   developing the targeted application schedule from the predicted below-ground performance for an application of a specific soil-active agricultural biological product at desired time or at a desired geographical location, wherein a user applies the specific soil-active agricultural biological product based on the targeted application schedule at the desired time or at the desired geographical location, or an automated application of the specific soil-active agricultural biological product at the desired time or at the desired geographical location is controlled based on the targeted application schedule.   
     
     
         12 . The method of  claim 11 , further comprising applying the environmental profile to provide one or more advisories regarding an application of the soil-active agricultural biological product in any field at any selected time based on the prediction of the performance of the soil-active agricultural biological product. 
     
     
         13 . The method of  claim 11 , further comprising analyzing a plurality of input data relative to the soil system, the particular field, and the soil-active agricultural biological product, the input data including soil-water information for the particular field, the soil-water information including one or more of surface and sub-surface drainage characteristics, an amount of water applied to the soil system through effective rainfall and irrigation, an amount of water that is retained in a dynamically-growing root zone, an amount of water depleted from the dynamically-growing root zone, and irrigation-specific information relative to past or recent irrigation activity. 
     
     
         14 . The method of  claim 13 , wherein the input data further includes field-specific data that includes positional information identifying the particular field, the positional information including GPS data that enables usage of GPS receivers to determine one or more of positional coordinates and boundaries of the particular field 
     
     
         15 . The method of  claim 13 , wherein in the input data further includes soil-specific information for the particular field, the soil-specific information including one or more of a soil type, and known soil temperature and moisture content for at least one of a current growing season and a period of time extending up to and including a current growing season. 
     
     
         16 . The method of  claim 13 , wherein in the input data further includes crop-specific information for the particular field, the crop-specific information including one or more of crop type data, planting data, crop growth data, growth stage-dependent parameters indicative of one or more crop growth stages, crop relative maturity data, crop planting depth and row spacing data, and targeted crop moisture or temperature thresholds. 
     
     
         17 . The method of  claim 13 , wherein the input data further includes field-level remotely-sensed imagery data of the particular field. 
     
     
         18 . The method of  claim 11 , further comprising developing one or more geographically-dispersed machine learning base models for a particular soil-active agricultural biological product, and applying the one or more geographically-dispersed machine learning base models to environmentally match the particular soil-active agricultural biological product for a particular geographic area. 
     
     
         19 . The method of  claim 18 , further comprising training the one or more geographically-dispersed machine learning base models with performance data representing efficacy of the particular soil-active biological catalyst for a smaller local region to re-learn the one or more geographically-dispersed machine learning base models for a specific localized geographical area within the smaller local region. 
     
     
         20 . The method of  claim 11 , further comprising building a comprehensive dataset for the customized field model of one or more physical and empirical characteristics impacting the soil system to automatically develop an artificial intelligence model to analyze time-varying characteristics of the soil system, and applying the artificial intelligence model to predict a below-ground performance of a soil-active agricultural biological product in any field at any selected time. 
     
     
         21 . A system, comprising:
 a computing environment including at least one computer-readable storage medium having program instructions stored therein and a computer processor operable to execute the program instructions to model a below-ground performance of a soil-active agricultural biological product within a plurality of data processing modules, the plurality of data processing modules including:   an initialization module configured to ingest above-ground weather data, represented by weather and climatological information that includes at least one of recent and current field-level weather data, historical field-level weather data and extended-range weather forecast data for a specific geographical area, crop-specific information relative to a crop in a particular field in the specific geographical area, soil-specific information relative to a soil system of the particular field, and a measured efficacy of the soil-active agricultural biological product in an outcome of one or more prior application of the soil-active agricultural biological product, wherein the soil-active agricultural biological product includes at least one of a bio-pesticide, a bio-stimulant, and a plant growth regulator;   one or more modules configured to develop a customized field model of one or more physical and empirical characteristics impacting the below-ground performance of the soil-active agricultural biological product in a soil system in a particular field, to enable a targeted application schedule at a desired time and a desired place to obtain a specific outcome, the customized field model configured to assess below-ground spatial and temporal soil system attributes that at least include soil moisture and soil temperature at multiple depths and over multiple time periods, and with a specific soil-active agricultural biological product applied to the soil system in the particular field and characterize environmental factors influencing performance of the soil-active agricultural biological product in the soil system of the particular field, and develop an environmental profile for   the performance of the soil-active agricultural biological product at a below-surface state represented in the below-ground spatial and temporal soil system attributes, by identifying one or more appropriate environmental predictors impacting biological processes in the particular field to analyze the below-surface state of the soil system in one or more machine learning models configured to construct variables describing the below-surface state and summarizing a correlated significance of above-ground weather data for a specific geographical area including the particular field and the below-ground spatial and temporal soil system attributes, and match the variables summarizing the correlated significance with the one or more prior applications of the soil-active agricultural biological product;   a prediction module configured to generate a prediction of the below-ground performance of the soil-active agricultural biological product from the environmental profile; and   an output module configured to develop the targeted application schedule from the predicted below-ground performance for application of a specific soil-active agricultural biological product at desired time or at a desired geographical location, wherein a user applies the specific soil-active agricultural biological product based on the targeted application schedule at the desired time or at the desired geographical location, or an automated application of the specific soil-active agricultural biological product at the desired time or at the desired geographical location is controlled based on the targeted application schedule.   
     
     
         22 . The system of  claim 21 , wherein the environmental profile is applied to provide one or more advisories regarding an application of the soil-active agricultural biological product in any field at any selected time based on the prediction of the performance of the soil-active agricultural biological product. 
     
     
         23 . The system of  claim 21 , wherein the input data further includes field-specific data that includes positional information identifying the particular field, the positional information including GPS data that enables usage of GPS receivers to determine one or more of positional coordinates and boundaries of the particular field. 
     
     
         24 . The system of  claim 21 , wherein the input data further includes soil-water information for the particular field, the soil-water information including one or more of surface and sub-surface drainage characteristics, an amount of water applied to the soil system through effective rainfall and irrigation, an amount of water that is retained in a dynamically-growing root zone, an amount of water depleted from the dynamically-growing root zone, and irrigation-specific information relative to past or recent irrigation activity for the particular field. 
     
     
         25 . The system of  claim 21 , where the soil-specific information includes one or more of a soil type, and known soil temperature and moisture content for at least one of a current growing season and a period of time extending up to and including a current growing season. 
     
     
         26 . The system of  claim 21 , wherein the crop-specific information further includes one or more of crop type data, planting data, crop growth data, growth stage-dependent parameters indicative of one or more crop growth stages, crop relative maturity data, crop planting depth and row spacing data, and targeted crop moisture or temperature thresholds. 
     
     
         27 . The system of  claim 21 , wherein the input data further includes field-level remotely-sensed imagery data of the particular field. 
     
     
         28 . The system of  claim 21 , wherein the one or more modules are further configured to develop one or more geographically-dispersed machine learning base models for a particular soil-active agricultural biological product, and apply the one or more geographically-dispersed machine learning base models to environmentally match the particular soil-active agricultural biological product for a particular geographic area. 
     
     
         29 . The system of  claim 28 , wherein the one or more modules are further configured to train the one or more geographically-dispersed machine learning base models with performance data representing efficacy of the particular soil-active agricultural biological product for a smaller local region to re-learn the one or more geographically-dispersed machine learning base models for a specific localized geographical area within the smaller local region. 
     
     
         30 . The system of  claim 21 , wherein the one or more modules are further configured to build a comprehensive dataset for the customized field model of one or more physical and empirical characteristics impacting the soil system to automatically develop an artificial intelligence model to analyze time-varying characteristics of the soil system, and applying the artificial intelligence model to predict a below-ground performance of a soil-active agricultural biological product in any field at any selected time.