Iterative forecasting using process-models

This section builds on the concepts in the book to demonstrate how to develop forecasts using process-based ecosystem models. First, the model is introduced, the key components of its uncertainty are examined, and data are obtained for calibration. Second, the model is calibrated using a Bayesian approach and initial conditions are determined using a particle filter. Finally, the forecast is iteratively submitted to the NEON Ecological Forecasting Challenge and past forecasts submitted to the challenge are evaluated. It is designed to be a practical guide for similar forecasting applications using other process models. The final chapter provides an example problem set for applying the skills in different contexts.