g in 
for 
| on 
ing, 
AN ECOLOGICAL REGIONALIZATION MODEL BASED ON NOAA/AVHRR DATA 
Yingchun Zhou 
Department of Geography, University of Nebraska-Lincoln, Lincoln, NE 68508, USA 
Email: yzhou@unlinfo.unl.edu 
Commission IV, Working Group 1 
KEY WORDS: Modelling Remote, Sensing GIS Resource Management Vegetation Ecoregion Mapping 
ABSTRACT 
Ecoregions are developed to assist in natural resource management and policy making. A quantitative, multivariate 
regionalization model was developed and applied on vegetation region mapping of Nebraska, the United States. The 
model aggregated small ecosystems into larger regions in a hierarchical clustering procedure. Ecosystem elements were 
merged into clusters based on the similarities of their ecological features, and their spatial neighborhood derived from 
GIS topological files. Vegetation regions of Nebraska were generated using digital State Soil Geographical data and 
multitemporal NOAA/AVHRR NDVI data of the growing season in 1991. 
INTRODUCTION 
Ecoregions are geographical zones that contain a number 
of similarly functioning ecosytems. They may represent 
broad similarities in ecosystem components including 
climate, geology, geomorphology, soils, and vegetation 
(Bailey et al, 1985; Wiken, 1986), or they may be 
designed to address more specific themes, such as 
vegetation types, soils, or water quality. Ecoregion 
frameworks are developed to assist in natural resource 
and environment management and policy making. In the 
United States, several ecoregion schemes have been 
developed. Kuchler (1970) mapped potential natural 
vegetation of the Unite States. The Land Resource 
Regions and Major Land Resource Areas (MLRA) were 
developed by the U.S. Department of Agriculture (USDA) 
to assist agricultural management (USDA, 1981). Other 
major frameworks are Ecoregions of the United States by 
the USDA Forest Service, and Ecoregion Maps compiled 
by the U.S. Environmental Protection Agency (Gallant et 
al, 1989). Ecoregions provide a framework in which 
similar responses may be expected within relatively 
homogeneous areas. Therefore, it is possible to formulate 
management policy and apply it on a regionwide basis 
rather than a site-by-site basis (Bailey et al, 1985). 
Most previous ecological regionalizations have been 
completed by a qualitative approach which employs 
continual, interactive expert judgements for selecting, 
analyzing, and classifying data in order to generate regions 
(Gallant et al, 1989). Quantitative tools are not sufficiently 
developed for incorporating the multivariate judgements 
needed to delineate regions. The objectives of this 
research were to develop a quantitative, multivariate 
regionalization model and apply it onto satellite imagery 
for vegetation zoning in Nebraska of the United States. 
1001 
The model was developed to aggregate small ecosystems 
into large ecoregions based on similarity of biotic and 
abiotic features. The mergers of ecosystems can be 
conceptualized by using a hierarchical dendrogram. In 
this research, State Soil Geographic (STATSGO) data map 
units were defined as the elementary ecosystems.  Bi- 
weekly composites of Normalized Difference Vegetation 
Index (NDVI) derived from NOAA/AVHRR were used as 
surrogates of vegetation conditions. 
STUDY AREA AND DATA SOURCES 
The state of Nebraska lies between 40° and 43° north 
latitude, and between 95.5° and 104° west longitude 
(Figure 1). Nebraska is part of the broad region which 
gently slopes up from the Mississippi River in the east to 
the Rocky Mountains in the west. The sedimentary rocks 
of this area are of several types, including limestone, 
sandstone, and shale. Located in the interior of North 
America, Nebraska has a continental climate with a 
considerable temperature range from summer to winter. 
The total growing season precipitation is generally 
adequate for crop production (Searcy and Longwell, 1964). 
Nebraska is primarily a grassland area with some forests 
and woodlands distributed along river valleys throughout 
the state. Most of its land has been cultivated for 
agricultural use. 
NOAA satellites provide daily Advanced Very High 
Resolution Radiometer (AVHRR) data at a ground 
resolution of 1x1 km. The coarse resolution and high 
repeat frequency of AVHRR data facilitate large area study 
and environmental monitoring. In vegetation mapping, it 
is possible to create phenological profiles from 
multitemporal AVHRR data to discriminate vegetation 
International Archives of Photogrammetry and Remote Sensing. Vol. XXXI, Part B4. Vienna 1996