Community Forecasting

Time-series with year on x and Dipodomys Merriami abundance on y. Time-series oscillates with significant temporal autocorrelation over 20 years. At the end of the time-series a forecast is shown with a large prediction internal.

We are broadly interested in forecasting the future state of ecological systems, both to provide actionable information about how these systems will change and to assess how well models capture the key information about how ecological systems operate. As part of this effort we are makign forecasts in automated iterative ways so that the forecasts are automatically updated and evaluated as new data is collected. We are using these methods to understand the impact of biological context on forecasting using the long-term experiments that are part of the Portal Project