Search Results

You are looking at 1 - 2 of 2 items for :

  • Author or Editor: David A. Marks x
  • Earth Interactions x
  • All content x
Clear All Modify Search
T. Scott Rupp, Xi Chen, Mark Olson, and A. David McGuire


Projected climatic warming has direct implications for future disturbance regimes, particularly fire-dominated ecosystems at high latitudes, where climate warming is expected to be most dramatic. It is important to ascertain the potential range of climate change impacts on terrestrial ecosystems, which is relevant to making projections of the response of the Earth system and to decisions by policymakers and land managers. Computer simulation models that explicitly model climate–fire relationships represent an important research tool for understanding and projecting future relationships. Retrospective model analyses of ecological models are important for evaluating how to effectively couple ecological models of fire dynamics with climate system models. This paper uses a transient landscape-level model of vegetation dynamics, Alaskan Frame-based Ecosystem Code (ALFRESCO), to evaluate the influence of different driving datasets of climate on simulation results. Our analysis included the use of climate data based on first-order weather station observations from the Climate Research Unit (CRU), a statistical reanalysis from the NCEP–NCAR reanalysis project (NCEP), and the fifth-generation Pennsylvania State University–NCAR Mesoscale Model (MM5). Model simulations of annual area burned for Alaska and western Canada were compared to historical fire activity (1950–2000). ALFRESCO was only able to generate reasonable simulation results when driven by the CRU climate data. Simulations driven by the NCEP and MM5 climate data produced almost no annual area burned because of substantially colder and wetter growing seasons (May–September) in comparison with the CRU climate data. The results of this study identify the importance of conducting retrospective analyses prior to coupling ecological models of fire dynamics with climate system models. The authors’ suggestion is to develop coupling methodologies that involve the use of anomalies from future climate model simulations to alter the climate data of more trusted historical climate datasets.

Full access
Eraldo A. T. Matricardi, David L. Skole, Mark A. Cochrane, Jiaguo Qi, and Walter Chomentowski


Selective logging degrades tropical forests. Logging operations vary in timing, location, and intensity. Evidence of this land use is rapidly obscured by forest regeneration and ongoing deforestation. A detailed study of selective logging operations was conducted near Sinop, State of Mato Grosso, Brazil, one of the key Amazonian logging centers. An 11-yr series of annual Lansdat images (1992–2002) was used to detect and track logged forests across the landscape. A semiautomated method was applied and compared to both visual interpretation and field data. Although visual detection provided precise delineation of some logged areas, it missed many areas. The semiautomated technique provided the best estimates of logging extent that are largely independent of potential user bias. Multitemporal analyses allowed the authors to analyze the annual variations in logging and deforestation, as well as the interaction between them. It is shown that, because of both rapid regrowth and deforestation, evidence of logging activities often disappeared within 1–3 yr. During the 1992–2002 interval, a total of 11 449 km2 of forest was selectively logged. Around 17% of these logged forests had been deforested by 2002. An intra-annual analysis was also conducted using four images spread over a single year. Nearly 3% of logged forests were rapidly deforested during the year in which logging occurred, indicating that even annual monitoring will underestimate logging extent. Great care will need to be taken when inferring logging rates from observations greater than a year apart because of the partial detection of previous years of logging activity.

Full access