Skip to main content
ARS Home » Crop Production and Protection » Research » Research Project #444375

Research Project: Climate Data, Modeling Support and Delivery for NPMTI

Location: Crop Production and Protection

Project Number: 0500-00102-001-028-S
Project Type: Non-Assistance Cooperative Agreement

Start Date: Jun 1, 2023
End Date: Jun 30, 2024

Objective:
The main objective of the Climate Data, Modeling Support and Delivery for the National Predictive Modeling Tool Initiative project is to give collaborators a single point for access to historical and forecasted climate data, uploading and downloading of field observations, and an automated process for developing predictive disease models utilizing machine learning (ML) based on those climate and observation data using python and/or R programming languages. These data and models will then be available for viewing in a dashboard environment with customizable visualizations.

Approach:
The collaborator will create a unified database of climate data that other collaborators within the National Predictive Modeling Tool Initiative will have access to. In addition, a dashboard will be created that allows viewing of weather data as well as filtering capabilities prior to download so that only data of interest is acquired. Researchers’ own data can also be loaded into the collaborator's platform, enabling them to view it in the context of other data, generate models using their data via python and R programming languages, and have it securely stored on a cloud-based Azure instance. In addition to this, the Collaborator will provide frameworks that other collaborators can use to generate models using a combination of their own data and Collaborator-provided climate data and environments. This will be accomplished through utilizing a web-based interactive computing platform that is leveraged to create an environment for researchers to create, test, and validate models utilizing Python and R programming languages. This will enable a very flexible solution allowing rapid development and validation of new models. All data and models will be stored on a cloud hosting services using industry-standard security protocols and measures.