Anthony Marcozzi Presents at 4th Southwest Fire Ecology Conference
Anthony Marcozzi Presents at 4th Southwest Fire Ecology Conference
Anthony Marcozzi, a fire scientist with the New Mexico Consortium and the Center for Applied Fire and Ecosystem Science, recently gave a talk titled Spatial inventory distribution in FastFuels with semantic segmentation and imputation at the 4th Southwest Fire Ecology Conference.
The Southwest Fire Ecology Conference was held November 18-22, 2024 in Santa Fe, New Mexico. The goal of this conference was to bring together professionals to share knowledge, exchange ideas, and discuss the latest advancements in fire ecology research and management with a focus on the southwestern United States.
Marcozzi’s presentation is a joint project with collaborators: Jacob Downs of University of Montana, Russel Parsons from the USFS, Jesse Johnson of the University of Montana, Lucas Wells, Daihti Martin, and Niko Tutland of the New Mexico Consortium
FastFuels is a novel data platform for creating and editing 3D fuels data with a primary objective of providing inputs to next generation fire models such as QUIC-Fire. To achieve this objective, tree inventory data is a critical data source. Currently, the FastFuels system uses the TreeMap raster product in conjunction with the Forest Inventory and Analysis (FIA) Database to achieve national coverage tree inventory data. However, the TreeMap raster product provides data at 30-meter resolution, does not provide plot-level spatial information, and does not consider multi-conditioned FIA plots.
In this presentation, Marcozzi proposed a novel methodology for leveraging high resolution canopy height models (CHMs), either from LIDAR acquisitions or machine learning inference, to semantically segment forest cover. Next, we propose a modified imputation technique which aims to improve the spatial resolution of the existing TreeMap product. He also proposed a crown-volume guided poisson disc sampling algorithm to spatially distribute tree inventory data by using a surface cost minimizing technique. This new modeling approach aims to provide large scale tree inventory coverage with special focus on the accuracy of tree locations and stand density in order to improve prescribed fire modeling applications.
Marcozzi states, “Bridging the gap between research and management in the form of new fire behavior models helps reduce wildfire risk, improve fuel reduction treatments, and put more good fire on the landscape. This is a big piece of addressing the wildfire crisis and NMC tools and research are a critical part of the solution”.
To learn more about the 4th Southwest Fire Ecology Conference see the webpage at: https://swfireconference.org/
To learn more about the Center for Applied Fire and Ecosystem Science and Anthony Marcozzi, see the CAFES webpage at: https://appliedfirescience.org/