Ecological Forecasting in Practice: Building Iterative Workflows for Biodiversity Predictions

Day 1. Introduction to Ecological Forecasting

• The role of forecasting in ecology and conservation
• From explanation to prediction: forecasting frameworks
• Near-term iterative forecasting vs long-term projections
• Overview of forecasting workflows (data → model → forecast → evaluation → update)
• Success versus quality in ecological forecasts
• Introduction to the EFFI forecasting challenges

Practical

• Course setup (R environment, data, repositories)
• Exploration of ecological and remote sensing datasets

Day 2. Working with Ecological and Remote Sensing Data

• Ecological time series: structure and challenges of integrating heterogeneous data
• Remote sensing for forecasting: Presenting most commonly used data types
• Linking satellite data with field observations (example: NDVI)
• Data preprocessing and feature extraction

Practical

• Processing NDVI data (cleaning, smoothing, deriving metrics)
• Exploring seasonal dynamics and trends

Day 3. Building Forecast Models

• Introduction to forecasting models
  • Autoregressive models (Hierarchical Bayesian models)
  • Correlational models (GLMMs)
  • Mechanistic models (incorporating physiology and species interactions))
• Model assumptions and limitations
• Iterative forecasting concepts

Practical

• Fitting forecasting models in R
• Generating short-term forecasts of biodiversity dynamics

Day 4. Uncertainty, Evaluation, and Reproducible Workflows

• Sources of uncertainty in ecological forecasts
• Representing and propagating uncertainty
• Forecast evaluation and model comparison
  • Metrics (e.g., Mean Squared Error)
  • Benchmark models (e.g., persistence, climatology)
• Introduction to reproducible forecasting workflows
  • Version control (Git)
  • Environment management (renv / Docker concepts)

Practical

• Evaluating forecast performance
• Adapting and improving models based on results

Day 5. Forecasting Challenge and Applications

• Forecasting challenges as a tool for scientific advancement
• Standardized forecast formats and submission
• Linking forecasts to decision-making and adaptive management

Practical

• Final model development
• Preparing and submitting forecasts

Discussion and wrap-up

• Comparison of model performance
• Strengths and limitations of different approaches
• Future directions in ecological forecasting