Spatial representation of relative flammability produced through summarization of the ALFRESCO model outputs. Learn more about these data
"Essentially, all models are wrong, but some are useful." — Statistician Greg E.P. Box
We use computer-generated models of physical and biological interactions to project (not predict) future environmental conditions. Projections describe what may happen to a system under the influence of assumptions such as higher temperatures or increases in wildfire.
Our models are useful for resource managers, community planners, businesses, and other institutions because they provide a way to think about the full extent of likely future environmental conditions. Models describe a range of possible futures that may happen under sets of assumptions known as scenarios. Their validity can be tested, and they provide a useful tool for managing uncertainty when planning for the future. Our modeling process includes:
- Collaboration We work with a large network of domestic and international researchers to improve our models and answer questions posed by our collaborators. Our home institution, the University of Alaska Fairbanks, is the center of Alaska and Arctic expertise.
- Data development We locate and develop input datasets, working on 10–100 year horizons and regional to global scales.
- Iteration Our computational capacity and flexibility allow us to rapidly process data and respond to collaborators’ needs.
- Validation We validate our models with historical observational data.
- Interpretation Our skilled staff and network of experts enables us to interpret, explain, and visualize projections of climate, ecosystem, and biophysical outputs.
Major ecosystem modeling efforts
The following models use SNAP climate data. SNAP and collaborators are also designing an Integrated Ecosystem Model that will link these models to help resource managers understand the nature and expected rate of landscape change.
ALFRESCO is a frame-based, spatially explicit model that shows spatial processes of fire and recruitment across the circumpolar arctic/boreal zone. It combines disturbance events, seed dispersal, and succession on a landscape at a spatio-temporal scale appropriate for investigating effects of climatic change.
A frame-based model partitions temporal changes in vegetation into states or frames. Each frame runs as an independent submodel, simulating processes important to a particular frame and which may cause a switch to a different frame. Variables (drivers) include:
- climate (growing-season temperature and precipitation)
- disturbance (fire)
- seed dispersal
ALFRESCO is based on the idea that vegetation state can be explained using probabilistic—or subject to chance—rules that respond to the model drivers of disturbance and climate. The model assumptions reflect a supposition that fire regime and climate are the primary drivers of landscape-level changes in the distribution of vegetation in this landscape. Learn more about SNAP's work with the ALFRESCO model in habitat research.
ALFRESCO operational details
ALFRESCO accepts SNAP climate data to project impacts of changing climate on fire regimes. It operates year-to-year in a landscape composed of 1 km × 1 km pixels (appropriate for communicating with mesoscale climate and carbon models). It uses vegetation types that simulate the complex mosaic typical of the Far North:
- Forest: black spruce, white spruce, deciduous
- Tundra: shrub, graminoid, wetland
- Barren lichen-moss
- Temperate rainforest
Rupp TS, Starﬁeld AM, Chapin FSI, Duffy P. 2002. Modeling the impact of black spruce on the ﬁre regime of Alaskan boreal forest. Climate Change 55:213–233.
Rupp TS, Starﬁeld AM, Chapin FS III. 2000. A frame-based spatially explicit model of subarctic vegetation response to climatic change: comparison with a point model. Landscape Ecology 15:383–400.
Yuan, F-M, Yi S-H, McGuire AD, Johnson KD, Liang J-J, Harden JW, Kasischke E, Kurz WA. 2012. Assessment of historical boreal forest carbon dynamics in the Yukon River Basin: Relative roles of climate warming and fire regime changes. Ecological Applications 22:2091-2109.
The TEM uses spatially referenced information on climate, elevation, soils, vegetation, and water availability to estimate carbon and nitrogen dynamics and pool sizes for Earth’s terrestrial ecosystems. Several versions of this model have been developed over the past decade thanks to improvements in computer resources, which have allowed more detailed studies of global ecosystem processes. Learn more about the TEM
The GIPL model simulates permafrost dynamics and the dynamics of thawing/freezing processes in the active layer (the depth of summer seasonal thaw in perennially frozen ground and depth of seasonal frost penetration in permafrost free areas) changes in soil temperature, and changes in permafrost extent.
Why is this important? Changes in permafrost can trigger substantive changes in hydrology, carbon cycling, and landscape structure, impacting both ecosystems and the built environment (infrastructure). Learn more about GIPL