M. Jabloun

The paper proposes a method to develop calibration guidelines for system models by analyzing multi-model simulations, applying it to crop model phenology calibration and providing recommendations on objective functions, parameters, and software.

This study evaluates wheat phenology prediction models in Australia, finding that multi-model averaging performs nearly as well as the best single model, with an average error of 9 days.

This study evaluates how well 27 crop modeling groups predict wheat phenology in Northern France using calibration data from the same target population, finding mean absolute errors comparable to measurement error.