Print Friendly and PDF


Advanced Search Results Detail

Project ID: 16-1-04-2

Year: 2016

Date Started: 08/01/2016

Ending Date:  07/31/2019

Title: Using multi-scale spatial data to improve predictions of immediate and delayed fire mortality

Project Proposal Abstract: Density-dependent mortality - the situation in which probability of mortality is higher in local areas of high tree density - is one of the most universal ecology concepts guiding forest management. Stand dynamics theory, numerous recent empirical tree mortality studies, including those of post-fire tree mortality, and the forest entomology literature all emphasize that local crowding increases the likelihood of tree death. Yet, the post-fire tree mortality equations used in common modeling platforms (e.g., FOFEM) do not include terms for local crowding, and density-dependence was not considered when those mortality models were developed. This raises the question as to whether current models are comprehensive and fully appropriate, especially in forests with heterogeneous tree densities or heterogeneous burn severities. Mangers need models that work across the spectrum of pre-fire and post-fire conditions to credibly guide decisions about post-fire landscape management, highlighting the need to rigorously evaluate current models. We will evaluate existing models of post-fire tree mortality to determine the potential improvement in accuracy offered by incorporating information about local tree neighborhood density. Despite the increasing magnitude of tree mortality and the importance of post-fire refugia for forest regeneration, the spatial distribution of fire mortality remains understudied. When spatial patterns of fire mortality are considered, they are usually inferred from Landsat-derived spectral changes. While landscape scale fire mortality has been inferred from these satellite-derived metrics, fire severity is approximated using correlations with ground measurements that include no explicit spatial information (i.e., Composite Burn Index plots) and broad categories of trees rather than distinct tree species and sizes. And despite the wide use of satellite-derived spectral indices, no one has yet quantified the precise tree death that is associated with Landsat-derived spectral changes. The ambiguities in interpreting tree mortality from spectral changes between pre-fire and post-fire images, including the various combinations of individual tree mortality that can lead to similar spectral changes, and the sensitivities of Landsat-derived fire severity values to location uncertainties of Landsat pixels remain serious barriers to more effective use of available satellite data (e.g., Monitoring Trends in Burn Severity). The reason that spatially explicit post-fire mortality has been understudied is the requirement for a study site of sufficient size where the trees and fuel have been mapped pre-fire, where the data are not lost during the fire, and where measurements of tree mortality are repeated post-fire. Examining spatial neighborhoods of tree mortality requires large, mapped plots, almost certainly >1 ha, with sizes >10 ha more likely to produce meaningful results. Furthermore, to infer statistically significant relationships between tree mortality and Landsat data would require hundreds of contiguous Landsat pixels within which the trees were all mapped. We propose to address these issues with a unique, matched set of pre-fire and post-fire data - a 25.6 ha plot (800 m × 320 m) where all 34,579 trees have been mapped and tracked from three years before the Rim Fire to two years post-fire, and will be tracked over to five years post-fire. This plot contains 238 contiguous Landsat pixels and experienced 24,319 immediate fire mortalities, with 7,408 trees surviving two years post-fire. This unparalleled field dataset, collected to specifically examine the spatial scales of fire effects, is enhanced by matched high resolution LiDAR data collected pre-fire, providing opportunities to extrapolate our results over large landscapes. This study site offers perhaps the best opportunity in the world today to examine the spatial aspects of prompt and delayed fire mortality and evaluate existing tree mortality models.

Principal Investigator: James A. Lutz

Agency/Organization: Utah State University

Branch or Dept: Department of Wildland Resources


Other Project Collaborators

Type

Name

Agency/Organization

Branch or Dept

Agreements Contact

Kellie S. Hedin

Utah State University

Sponsored Programs

Budget Contact

Kellie S. Hedin

Utah State University

Sponsored Programs

Co-Principal Investigator

Van R. Kane

University of Washington

School of Environmental and Forest Sciences

Co-Principal Investigator

Andrew J. Larson

University of Montana

College of Forestry & Conservation


Project Locations

Fire Science Exchange Network

California


Level

State

Agency

Unit

STATE

CA

NPS

Yosemite National Park


Project Deliverables

There is no final report available for this project.
There are no deliverables available for this project.

Supporting Documents

There are no supporting documents available for this project.

Convert PDF documents to an html document using Adobe's online conversion tool.
Download Adobe Acrobat Reader