Guidance for forest management and landscape ecology applications of recent gradient nearest neighbor imputation maps in California, Oregon, and Washington / David M. Bell [and three others].

Format:

Book

Government document

Contributor:

Bell, David M.

Gregory, Matthew J.

Palmer, Marin

Davis, Raymond

Pacific Northwest Research Station (Portland, Or.)

Series:

General technical report PNW ; 1018.

General technical report PNW ; GTR-1018

Language:

English

Subjects (All):

Forest management--Pacific States.

Forest management.

Landscape ecology--Pacific States.

Landscape ecology.

Forest mapping--Pacific States.

Forest mapping.

Nearest neighbor analysis (Statistics).

United States--Pacific States.

Physical Description:

1 online publication (41 pages) : illustrations, maps.

Place of Publication:

Portland, OR : United States Department of Agriculture, Forest Service, Pacific Northwest Research Station, 2023.

Summary:

"For many forest landscape ecology and ecological monitoring projects, forest structure and composition data availability at the correct scale are often limiting factors, motivating the development of map products based on remote sensing. Gradient nearest neighbor (GNN) imputation is a flexible framework for generating multivariate, annual, wall-to-wall maps of forest structure and composition for landscape, regional, and national applications. This report provides guidance on the appropriate use of forest structure and composition maps generated from satellite imagery, physical environment, and forest inventory data using the GNN modeling and mapping framework. We describe the GNN modeling and mapping framework associated with the generation and delivery of updated maps in 2020 (GNN-2020) that provide forest attribute status and trend data from 1986 to 2017. In relation to the GNN-2020 map data, we describe (1) the accuracy assessment reporting that accompanies all maps, (2) basic concepts regarding the strength of relationships between forest attributes and the geospatial predictor variables used in mapping, (3) the role of the number of nearest neighbors in map accuracy, (4) factors affecting the appropriate spatial and temporal scales for using the maps, (5) the types of changes in forest attributes that can be reasonably assessed with the maps, and (6) the challenges in creating categorical or classified maps based on the GNN-2020 data. Our goal is to provide enough background and guidance to use GNN products appropriately and effectively."

Notes:

"November 2023."

Includes bibliographical references (pages 31-41).

OCLC:

1411843077