A Multi-Index Integrated Change detection method for updating the National Land Cover Database

By: , and 



Land cover change is typically captured by comparing two or more dates of imagery and associating spectral change with true thematic change. A new change detection method, Multi-Index Integrated Change (MIIC), has been developed to capture a full range of land cover disturbance patterns for updating the National Land Cover Database (NLCD). Specific indices typically specialize in identifying only certain types of disturbances; for example, the Normalized Burn Ratio (NBR) has been widely used for monitoring fire disturbance. Recognizing the potential complementary nature of multiple indices, we integrated four indices into one model to more accurately detect true change between two NLCD time periods. The four indices are NBR, Normalized Difference Vegetation Index (NDVI), Change Vector (CV), and a newly developed index called the Relative Change Vector (RCV). The model is designed to provide both change location and change direction (e.g. biomass increase or biomass decrease). The integrated change model has been tested on five image pairs from different regions exhibiting a variety of disturbance types. Compared with a simple change vector method, MIIC can better capture the desired change without introducing additional commission errors. The model is particularly accurate at detecting forest disturbances, such as forest harvest, forest fire, and forest regeneration. Agreement between the initial change map areas derived from MIIC and the retained final land cover type change areas will be showcased from the pilot test sites.

Publication type Conference Paper
Publication Subtype Conference Paper
Title A Multi-Index Integrated Change detection method for updating the National Land Cover Database
Year Published 2010
Language English
Publisher American Geophysical Union
Contributing office(s) Earth Resources Observation and Science (EROS) Center
Description GC51F-0804
Conference Title American Geophysical Union 2010 Fall Meeting
Conference Location San Francisco, CA
Conference Date December 13-17, 2010
Online Only (Y/N) N
Additional Online Files (Y/N) N
Google Analytic Metrics Metrics page
Additional publication details