Statistical Evaluation of Two Global Climate Datasets in Support of Pest Forecasting Models

Invasive pests pose a risk to agriculture and natural resources in the United States, causing economic and environmental harm and creating challenges for the trade of agricultural products both domestically and globally. Pest forecast models are utilized to help protect against the entry, establishment, and further spread of significant pests nationwide while also facilitating safe trading systems. These models rely on global climate datasets of temperature and precipitation to provide atmospheric conditions at a variety of spatial and temporal resolutions.

Results from this study provide global climate data recommendations for the models used in the Spatial Analytical Framework for Advanced Risk Information Systems (SAFARIS) project – a collaboration between the Center for Integrated Pest Management (CIPM) at NC State University and the United States Department of Agriculture (USDA) Animal and Plant Health Inspection Service (APHIS). The goal of SAFARIS is to contribute forecasts of the behavior of potentially harmful pests to APHIS’ Plant Protection and Quarantine (PPQ) program. These pest forecasts inform risk assessment and management for the United States and the entire globe.

This study quantitatively analyzes minimum and maximum temperature as well as precipitation data for two global climate datasets – Climate Forecast System Reanalysis (CFSR) and Climate Forecast System Version 2 (CFSv2) – to determine the level of accuracy as compared to Global Historical Climatology Network-Monthly (GHCN-M) data. Global error and bias of the two Climate Forecast System datasets are calculated using ground-station – grid-point pairs all over the world. In addition, a regional scale analysis is performed for each continent and sub-region within the continental United States.