Near-term ecological forecasting is an emerging imperative for supporting environmental decision making and adaptive management. However, creating automatic forecasts and accessing interpretable results can be a challenge for both researchers and stakeholders alike. Within the Near Term Ecological Forecasting Initiative (NEFI), we created a server to make it easier for researchers to implement, archive, and share iterative forecasts by providing the computation tools and framework required for running automatic forecasts. An important component of the system is a ShinyApp that displays interactive forecast results and model performance metrics. The NEFI system is generalized to run and visualize a wide range of forecasts and allows users to create public and private forecasts with public forecasts being available to all users through the ShinyApp. This generalized system for interactive, with interpretable visualizations is available to the public, stakeholders, managers, and researchers. Updated results are available daily and assist with decision support and allow for the advancement of near term forecasting.
Multiple forecasts have been deployed through the system and generalization of the visualizations app has been tested. Creating generalized yet informative visualizations for all forecasts in the systems provides challenges that we addressed through testing and user feedback. The system will be tested in June by NEFI summer course users, and participants will submit feedback on visualizations, ease of use, and interpretability. Additionally, we will request feedback from beta users when the system goes live in the summer to better meet the needs of users.
Results/Conclusions
Here we report our methods for creating the ShinyApp, present multiple forecast visualizations results (vegetation phenology, land-surface fluxes, soil microbes, ticks), and discuss feedback and results from user testing. NEFI strives to make forecasting more accessible for all researchers and creating generalized, interactive visualizations is an important component to effectively creating forecasts that are useful for the scientist, stakeholders, and managers.