Pest occurrence risk assessment and prediction in neighboring fields, crops and soils using crowd-sourced occurrence data

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 profile a risk assessment of pest occurrence within a plurality of data processing modules, the plurality of data processing modules including:
• an initialization module configured to 1) receive crowd-sourced reports that indicate a pest presence in a reporting field and identify a pest, 2) correlate positional coordinates of the reporting field and coordinates of one or more targeted fields from at least one GPS-tagged report in the crowd-sourced information from which a GPS receiver extracts the positional coordinates of the reporting field, and 3) identify and obtain additional input data relative to the pest presence, including one or more of pest data that at least includes biological characteristics of the pest, weather information that at least includes one or more of recent and current field-level weather data and short-term weather forecast data for both the reporting field and for the one or more targeted fields, and crop management data that includes at least one of crop planting dates and crop growth information over time for both reporting field and the one or more targeted fields;
• one or more modules configured to identify a set of descriptors relating to the pest presence in the reporting field from one or more correlated variable associations of the crowd-sourced observations and the additional input data to develop an expected pest-environment relationship in the reporting field, and construct an adaptive and localized infestation suitability model for the one or more targeted fields from the set of variables to perform a comparison between the reporting field and the one or more targeted fields from the expected pest-environment relationship, in an unsupervised learning model that 1) selects one or more environmental and crop management predictors from the set of descriptors based on a multivariate similarity among the correlated variable associations of the input data, 2) calculates statistical probabilities of similarity of pest occurrence relative to the one or more environmental and crop management predictors, and 3) continually updating the set of descriptors for the pest-environment relationship, the selection of the one or more of the environmental and crop management predictors based on the multivariate similarity among the correlated variable associations of the input data, and the statistical probabilities of similarity of pest occurrence as the crowd-sourced reports are received, and create a risk assessment profile for pest occurrence for the one or more targeted fields;
• and
• a pest occurrence prediction module configured to predict a pest occurrence from the risk assessment profile.

A method of assessing a field-specific risk of pest occurrence in a targeted field, comprising:

• identifying 1) a pest presence in a reporting field, and 2) and positional coordinates of the reporting field, from crowd-sourced information that includes at least one GPS-tagged report from which a GPS receiver extracts the positional coordinates of the reporting field for correlation of the positional coordinates of the reporting field with coordinates of one or more targeted fields;
• defining an expected pest-environment relationship for the reporting field by identifying a set of descriptors relating to the pest presence in the reporting field from one or more correlated variable associations of the crowd-sourced information, and one or more of pest data that at least includes biological characteristics of a pest identified in the reporting field, weather information that at least includes one or more of recent and current field-level weather data and short-term weather forecast data for the reporting field, and crop management data that includes at least one of crop planting dates and crop growth information over time for the reporting field;
• matching the expected pest-environment relationship for the reporting field at least with weather information that at least includes one or more of recent and current field-level weather data and short-term weather forecast data for the one or more targeted fields, and crop management data that at least one of crop planting dates and crop growth information over time for the one or more targeted fields, by constructing an adaptive and localized infestation suitability model for the one or more targeted fields for performing a comparison between the reporting field and the one or more targeted fields within an unsupervised learning engine, configured to 1) select one or more environmental and crop management predictors from the set of descriptors based on a multivariate similarity among the correlated variable associations, 2) calculate statistical probabilities of similarity of a pest occurrence relative to the selected one or more environmental and crop management predictors, and 3) continually updating the set of descriptors for the pest-environment relationship, selection of the one or more environmental and crop management predictors, and the statistical probabilities of similarity of a pest occurrence, as either of a pest presence in a reporting field or positional coordinates of the reporting field are identified from crowd-sourced information to profile a risk assessment of pest occurrence in the one or more targeted fields;
• and
• estimating a probability of pest occurrence for the one or more targeted fields from the profile of risk assessment.