You are here

Integrating ecosystem sampling, gradient modeling, remote sensing, and ecosystem simulation to create spatially explicit landscape inventories

Posted date: June 24, 2019
Publication Year: 
2002
Authors: Keane II, Robert E.; Rollins, Matthew G. G.; McNicoll, Cecilia H.; Parsons, Russell A.
Publication Series: 
General Technical Report (GTR)
Source: RMRS-GTR-92. Fort Collins, CO: U.S. Department of Agriculture, Forest Service, Rocky Mountain Research Station, 61 p.

Abstract

Presented is a prototype of the Landscape Ecosystem Inventory System (LEIS), a system for creating maps of important landscape characteristics for natural resource planning. This system uses gradient-based field inventories coupled with gradient modeling remote sensing, ecosystem simulation, and statistical analyses to derive spatial data layers required for ecosystem management. Field data were collected in two large (more than 10,000 km2) study areas along important environmental gradients using modified ECODATA methods. A multilevel database was used to derive response variables for predictive landscape mapping from the ECODATA database. Linkage of gradient models with remote sensing allows a standardized, flexible, detailed, and comprehensive classification of landscape characteristics. Over 40 spatially explicit variables were derived for each study area using existing spatial data, satellite imagery, and ecosystem simulation. This spatial database (the LEIS GIS) described landscape-scale indirect, direct, and resource gradients and provided predictor variables for multivariate predictive landscape models. Statistical programs and GIS were used to spatially model several landscape characteristics as a proof of concept for the LEIS. These proof-of-concept products were: (1) basal area, (2) western redcedar habitat, and (3) fuel models. Output maps were between 65 percent and 90 percent accurate when compared to reference data from each study area. Main strengths of the LEIS approach include: (1) a standardized, repeatable approach to sampling and database development for landscape assessment, (2) combining remote sensing, ecosystem simulation, and gradient modeling to create predictive landscape models, (3) flexibility in terms of potential maps generated from LEIS, and (4) the use of direct, resource, and functional gradient analysis for mapping landscape characteristics.

Citation

Keane, Robert E.; Rollins, Matthew G.; McNicoll, Cecilia H.; Parsons, Russell A. 2002. Integrating ecosystem sampling, gradient modeling, remote sensing, and ecosystem simulation to create spatially explicit landscape inventories. RMRS-GTR-92. Fort Collins, CO: U.S. Department of Agriculture, Forest Service, Rocky Mountain Research Station, 61 p.