Assessing simulated ecosystem processes for climate variability research at Glacier National Park, USA

Joseph D. White, Steven W. Running, Peter E. Thornton, Robert E. Keane, Kevin C. Ryan, Daniel B. Fagre, Carl H. Key

Research output: Contribution to journalArticlepeer-review

53 Scopus citations


Glacier National Park served as a test site for ecosystem analyses than involved a suite of integrated models embedded within a geographic information system. The goal of the exercise was to provide managers with maps that could illustrate probable shifts in vegetation, net primary production (NPP), and hydrologic responses associated with two selected climatic scenarios. The climatic scenarios were (a) a recent 12-yr record of weather data, and (b) a reconstituted set that sequentially introduced in repeated 3-yr intervals wetter-cooler, drier-warmer, and typical conditions. To extrapolate the implications of changes in ecosystem processes and resulting growth and distribution of vegetation and snowpack, the model incorporated geographic data. With underlying digital elevation maps, soil depth and texture, extrapolated climate, and current information on vegetation types and satellite-derived estimates of a leaf area indices, simulations were extended to envision how the park might look after 120 yr. The predictions of change included underlying processes affecting the availability of water and nitrogen. Considerable field data were acquired to compare with model predictions under current climatic conditions. In general, the integrated landscape models of ecosystem processes had good agreement with measured NPP, snowpack, and streamflow, but the exercise revealed the difficulty and necessity of averaging point measurements across landscapes to achieve comparable results with modeled values. Under the extremely variable climate scenario significant changes in vegetation composition and growth as well as hydrologic responses were predicted across the park. In particular, a general rise in both the upper and lower limits of treeline was predicted. These shifts would probably occur along with a variety of disturbances (fire, insect, and disease outbreaks) as predictions of physiological stress (water, nutrients, light) altered competitive relations and hydrologic responses. The use of integrated landscape models applied in this exercise should provide managers with insights into the underlying processes important in maintaining community structure, and at the same time, locate where changes on the landscape are most likely to occur.

Original languageEnglish
Pages (from-to)805-823
Number of pages19
JournalEcological Applications
Issue number3
StatePublished - Aug 1998


  • Climate variability
  • Ecological modeling
  • Glacier National Park
  • Model verification
  • Net primary productivity


Dive into the research topics of 'Assessing simulated ecosystem processes for climate variability research at Glacier National Park, USA'. Together they form a unique fingerprint.

Cite this