Latent trajectory models for spatio-temporal dynamics in Alaskan ecosystems.

Area of Science:

• Ecology
• Remote Sensing
• Climate Science

Background:

• The Alaskan Arctic is experiencing rapid environmental changes, including shrub and tree expansion.
• Understanding these ecosystem transformations is crucial for predicting future landscape dynamics.

Purpose of the Study:

• To quantify the impact of climate change on Arctic ecosystem structural transformation.
• To develop a predictive model for land cover transitions using remotely sensed data.

Main Methods:

• Developed a Bayesian hierarchical model with latent trajectory processes to analyze dynamic ecosystem states.
• Incorporated spatio-temporally heterogeneous climate drivers and accounted for temporal irregularity in survey data.
• Utilized a Pólya-Gamma sampling strategy for computational efficiency.

Main Results:

• The model successfully inferred rates of land cover transitions driven by climate change.
• Characterized multi-scale spatial correlations within the study system.
• Provided a framework for understanding ecosystem responses to climate shifts.

Conclusions:

• The developed model effectively quantifies climate change impacts on Arctic ecosystems.
• The model can predict future land cover changes under different climate scenarios.
• This research offers valuable insights into Arctic ecosystem dynamics and resilience.