LTER Sweet Cherry Model v1

The LTER–Sweet Cherry Model v1 is a web-based decision-support tool designed to forecast seasonal bud development and freeze risk in sweet cherry. The application integrates environmental data with physiological modeling to estimate dormancy progression, floral relative water content, and cold hardiness thresholds throughout the dormant season.

Using daily temperature and solar radiation inputs, the model generates seasonal predictions of: Dormancy progression using Low Temperature Exotherm Ratios (LTER), Relative water content (RWC) of floral buds and Cold hardiness thresholds including LT, LT10, and LT50.

Forecasts are available for 12 commercial sweet cherry cultivars as well as a species-level model. Interactive time-series visualizations allow growers, researchers, and industry professionals to interpret seasonal bud state and evaluate potential freeze risk by comparing predicted cold hardiness thresholds with observed minimum atmospheric temperature.

Model development utilized multi-season physiological datasets collected in Prosser, Washington (2019–2022) that included measurements of dormancy status, water relations, and cold hardiness aligned with daily atmospheric conditions. The modeling framework integrates multiple statistical approaches including GLM, random forest, neural networks, support vector machines, and gradient boosting models to generate seasonal forecasts with associated uncertainty bands.