Climate change and intensification of disturbance regimes are increasing the vulnerability of interior Alaska Department of Defense (DoD) training ranges to widespread changes in land cover and hydrology. This will have profound impacts on wildlife habitats, conservation objectives, permitting requirements, and military training activities.
The objective of this project was to provide U.S. Army Garrison Fort Wainwright Alaska (USAG-AK) training land managers a scientifically-based geospatial framework to assess wildlife habitat distribution and trajectories of change, and to identify vulnerable wildlife species whose habitats and resources are likely to decline in response to permafrost degradation, changing wildfire regimes, and hydrologic reorganization, projected to 2100.
Specifically, this project attempted to answer the question: “where, when, and how will projected climate warming in the area affect habitat access, suitability, and use?”
This project was focused on the following four specific objectives:
The project team linked an extensive program of field measurements, data synthesis, repeat imagery analyses, remote sensing measurements, and model simulations focused on land cover dynamics and wildlife habitat characteristics to identify suites of wildlife species most vulnerable to climate change.
The Alaska Thermokarst Model (ATM) is a state and transition probabilistic model that predicts change in ecotype distribution on an annual basis, at a 30-m resolution. Currently, the model represents the effect of wildfire, thermokarst, permafrost aggradation, river movement, succession, lake drainage, human activities (e.g., trails and clearings), shrubification and forest advancement on ecotype dynamic.
The ATM was coupled with two other models to better represent the biogeophysical processes driving the effect of projected climate change on landscape dynamics. The Alaska Frame Based Ecosystem Code was used to integrate spatially explicit simulations of fire occurrence and fire severity in response to multiple climate change scenarios. In addition, use was made of spatially explicit simulations of permafrost and vegetation dynamics from the Terrestrial Ecosystem Model to represent the effect of permafrost thaw and changes in soil moisture and organic layer thickness on land cover dynamics.
From this collected data, the project team created a robust database linking vegetation, soil, and environmental characteristics across interior Alaska training ranges.
In developing the database of all known measurements of permafrost presence/absence, soil type, and vegetation characteristics across the USAG-AK training land domain, the project identified many hundreds of measurements that the Army does not currently have in their geospatial datasets. These included measurements from a variety of studies that were not known or from contractors that had not finalized their datasets and/or had not shared them. Due to the remote nature of the training lands and the high time and monetary cost of these types of data, this expanded geospatial dataset has high value. There are other information sources still to be integrated but this is an exciting outcome both for the modeling needs (i.e., it provided more data than was known to exist), but also for the Army to have access to the broadest measurements available.
Conservation and protection of habitat conditions for the most vulnerable wildlife species of the study area helped address conservation objectives. Potentially vulnerable species discussed were bird and mammal habitat specialists, wood frog, rusty blackbird, and caribou.
The framework used was designed to support decision making for conservation management and habitat monitoring, land use, infrastructure development, and adaptive management across the interior Alaska DoD cantonment and training land domain.
The first runs of the ATM show promise in applying the model parameters in Interior Alaska. The comparison between modeled and observed rates of permafrost plateau loss shows strong statistically significant correlation. Based on current ATM outputs, there is a wide discrepancy in future wetland areal extent, which is driven by uncertainties in future greenhouse gas atmospheric concentration scenarios. The ecotype and habitat projections rely on assessment of wetland characteristics so this uncertainty in future scenario planning provides a stark reminder that a warming climate will be felt strongest in high latitudes.
Other significant results of the project include the following: