|the Art of Making Maps,
the Science of Where You Are
Maps of the Seattle Area
This image of the Seattle area was taken by the Multispectral
Scanner (MSS) on Landsat.
The satellite orbits the Earth at about 700 km altitude.
This is a single channel image: the brightness
of points in the image shows how much light of a specific
wavelength (color) the satellite sensor image recorded.
The MSS sensor has a spatial resolution of about 75 meters.
This means that, usually, you can't make out objects which are
smaller than about 75 meters.
Activity 1: What can you see?
- Using the small map above, find these features
in the Landsat image:
- follow Interstate-90 east from Seattle, across
Lake Washington, south of Lake Sammamish to the
right edge of the image.
- follow the Sammamish River as it flows from
the north end of Lake Sammamish (through an
agricultural area) to turn west and
flow into the north end of Lake Washington.
- As it crosses Lake Washington east of Mercer Island
(the large island in the south of the lake), I-90 is
12 lanes across.
Estimate the width of this bridge and convert to meters.
Is this less than, greater than, or about equal to the
- Do you or should you see I-90 in the image, or can
you only detect I-90 because of other related features?
If you detect I-90 because of related features, what are those
- The Sammamish River (sometimes called the Sammamish
Slough) flows in a dredged channel about 20 ft wide.
Convert this width to meters and compare to the Landsat
- Do you or should you see the Sammamish River in the image, or can
you only detect the river because of other related features?
If you detect the river because of related features, what are those
Down on the Corner...
View a larger image (over 1Mb!).
This is an aerial survey photo of the Lake Hills Greenbelt area
The photo is taken from an airplane flying at 20,000 ft.
The resolution in this photograph is about 1 meter.
The lake in the upper left is Larsen Lake; Phantom Lake is in the
(You can see both these lakes in the Landsat image at the top of
this page: they are north of I-90 and west of Lake Sammamish.)
This map of the Bellevue area includes the Greenbelt shown in the
Activity 2: Resolution and Altitude
How well you can see something depends on how close you are to it and
how good your vision is.
From across a room, you can probably tell a book from a box, but you
need to be closer to the book to be able to read its title.
Remote sensing systems are the same way.
The farther away they are, the more they can see, but with less detail.
Probably the easiest way to answer these questions is to use a graph.
Suppose you want to get an image with 20 m resolution.
What is a good altitude?
If you were at 500 km instead of 700 km, what kind of
resolution can you expect?
These questions might seem impossible to answer, but you can make
some simplifying assumptions to get a reasonable approximation.
- Convert 20,000 ft to km.
- Make a graph with altitude (in kilometers) on the x axis and
resolution (in meters) on the y axis.
- Plot the Landsat and aerial photograph information on this graph.
- Connect the two points with a straight line.
- Where do you get 20 m resolution?
(Find 20 m resolution on your y-axis, read across to the line
you drew, and go straight down to find the altitude which
gives you that resolution.
Since this is an approximation, report your answer to the nearest 100 km.)
- What resolution can you expect at 500 km?
(Find 500 km on the x-axis, follow it up to your line.
Read across to find the resolution you can get at this altitude.)
This is not an exact calculation, but often calculations like this help
scientists and engineers figure out if a job is going to be easy or
For example, knowing you can get 1 m resolution with a good
imaging system at 20,000 ft tells you that you need an extremely
good imaging system to get 1 m resolution at 700 km.
The basic assumptions you used for your approximation are:
1. The imaging systems on Landsat and the aerial survey missions
(This assumption lets you plot the aerial survey data and the Landsat
data on the same graph.)
2. "Resolution" refers to the width of an object, not its area.
(This assumption lets you draw a straight line between the points.)
This concludes our trip down from space. From here, you can
- take stock of what you've learned in
Cartography Review or
- bring your global view down to blades of grass
in a tour of the
Zoom back up to Puget Sound Maps.
Landsat image and aerial photograph -
USGS EROS Data Center
allows you to browse and select images.
Map of Seattle and vicinity from the Tiger Map Service of the US Census Bureau.
Detailed MapQuest! map of the Lake Hills Greenbelt area is copyright 1996 GeoSystems Global Corporation, Inc.
Hugh Anderson, and
Tell us what you think about this page.
Fri Jan 24 23:22:00 1997