landscapes and to compare ecological quality across the landscapes (O’Neill et al.,
1988; Ritters et al., 1995; McGarigal and Marks, 1995; Gustafson, 1998). Some
software has integrated landscape metrics to facilitate metrics calculation, for
example, FRAGSTATS (McGarigal et al., 2002).
Scale is an important issue in remote sensing and GIS studies. Scale influences the
examination of the landscape patterns in a region. Spatial scale is thought to affect the
quantification of landscape metrics (Turner, 1990; Wang et al., 1999). Choosing
different remote sensing sensors may cause various research results in the regional
analysis with diverse spatial resolutions. Because the spatial characteristics of large
regions on the level of the landscape have attracted researchers’ attentions (Franklin
and Forman, 1987; Turner, 1990), methods and technologies to examine the spatial
arrangements at broad spatial scales have become more and more important. Turner
(1990) developed a neutral model to examine the relationship between landscape
patterns and ecological processes across spatial scales. Kr€ onert et al. (2001) believed
that up/downscaling was a fundamental operation of transmission. They summarized
the classification of up/downscaling methods: Upscaling includes (i) averaging of
observations or output variables, (ii) finding representative parameters, (iii) averaging
of model equations, and finally (iv) model simplification. Downscaling includes (i)
empirical functions, (ii) mechanisitc models, and (iii) fine scale auxiliary information.
Scaling issue has been considered an important topic in landscape mapping. Imagery
with finer resolution includes greater spatial information, which, in turn, enables the
description of smaller features than imagery with lower spatial resolution. This scale
problem challenges the research of the relationship between spatial patterns and
processes (Meentemeyer, 1989).
Despite much research focused on human-related environment and related issues at
global and regional scales, some questions remain to be answered: What are the
methods to scale up or scale down across the landscape based on different data
resources? What are the scaling effects on the processes? We conducted a case study
to examine the changes in spatial configurations on landscape patterns and land
surface temperatures at different spatial resolutions. A logical approach was used to
examine the scaling-up effect on the measurements of LULC and LSTs by the use of
landscape metrics. The results helped to further understand the relationships between
landscape patterns and LSTs in urban environments.
11.2 CASE STUDY
11.2.1 Study Area
The city of Indianapolis, Indiana (Marion County, 39
47
0 N, 86
09
0 W) was selected as
study area in the case study (Figure 11.1). It is the nation’s 12th largest city and the
capital city of Indiana with a population of 0.8 million according to the U.S. Census
2010 (about 1.8 million in the metropolitan area). The city has clear seasonal changes,
but no pronounced wet or dry seasons. Its annual average temperature is 11.3
C, and
the average monthly temperature is 3.3
C in January and 23.9
C in July. The
218
SCALING ISSUES IN STUDYING THE RELATIONSHIP
1988; Ritters et al., 1995; McGarigal and Marks, 1995; Gustafson, 1998). Some
software has integrated landscape metrics to facilitate metrics calculation, for
example, FRAGSTATS (McGarigal et al., 2002).
Scale is an important issue in remote sensing and GIS studies. Scale influences the
examination of the landscape patterns in a region. Spatial scale is thought to affect the
quantification of landscape metrics (Turner, 1990; Wang et al., 1999). Choosing
different remote sensing sensors may cause various research results in the regional
analysis with diverse spatial resolutions. Because the spatial characteristics of large
regions on the level of the landscape have attracted researchers’ attentions (Franklin
and Forman, 1987; Turner, 1990), methods and technologies to examine the spatial
arrangements at broad spatial scales have become more and more important. Turner
(1990) developed a neutral model to examine the relationship between landscape
patterns and ecological processes across spatial scales. Kr€ onert et al. (2001) believed
that up/downscaling was a fundamental operation of transmission. They summarized
the classification of up/downscaling methods: Upscaling includes (i) averaging of
observations or output variables, (ii) finding representative parameters, (iii) averaging
of model equations, and finally (iv) model simplification. Downscaling includes (i)
empirical functions, (ii) mechanisitc models, and (iii) fine scale auxiliary information.
Scaling issue has been considered an important topic in landscape mapping. Imagery
with finer resolution includes greater spatial information, which, in turn, enables the
description of smaller features than imagery with lower spatial resolution. This scale
problem challenges the research of the relationship between spatial patterns and
processes (Meentemeyer, 1989).
Despite much research focused on human-related environment and related issues at
global and regional scales, some questions remain to be answered: What are the
methods to scale up or scale down across the landscape based on different data
resources? What are the scaling effects on the processes? We conducted a case study
to examine the changes in spatial configurations on landscape patterns and land
surface temperatures at different spatial resolutions. A logical approach was used to
examine the scaling-up effect on the measurements of LULC and LSTs by the use of
landscape metrics. The results helped to further understand the relationships between
landscape patterns and LSTs in urban environments.
11.2 CASE STUDY
11.2.1 Study Area
The city of Indianapolis, Indiana (Marion County, 39
47
0 N, 86
09
0 W) was selected as
study area in the case study (Figure 11.1). It is the nation’s 12th largest city and the
capital city of Indiana with a population of 0.8 million according to the U.S. Census
2010 (about 1.8 million in the metropolitan area). The city has clear seasonal changes,
but no pronounced wet or dry seasons. Its annual average temperature is 11.3
C, and
the average monthly temperature is 3.3
C in January and 23.9
C in July. The
218
SCALING ISSUES IN STUDYING THE RELATIONSHIP