Now we look at a few of the satellites and sensors commonly
used for remote sensing of Earth resources. These sensors
are all satellites oriented toward land-resources data.
They are either multispectral or panchromatic sensors.
Other satellites carry other kinds of sensors, and we will
examine a few of those in the next page. As you read this
page, you can refer to the chart below to help organize the
characteristics of each sensor.
| Satellite/ Sensor |
Landsat MSS |
Landsat TM |
SPOT XS |
SPOT Pan |
IRS 1-A/B LISS |
IRS 1-C Pan |
JERS-1 OPS |
NOAA AVHRR |
| Country of Origin | USA | USA | France | France | India | India | Japan | USA |
| Pixel Resolution (m) | 80 | 30 | 20 | 10 | 23 to 72 | 5.8 | 20 | 1000 |
| Number of bands | 4 | 7 | 3 | 1 | 4 | 1 | 7 | 5 |
| Panchromatic (B/W) | -- | -- | -- | 1 | -- | 1 | -- | -- |
| Blue | -- | 1 | -- | -- | 1 | -- | -- | -- |
| Green | 1 | 1 | 1 | -- | 1 | -- | 1 | -- |
| Red | 1 | 1 | 1 | -- | 1 | -- | 1 | 1 |
| Near/Mid IR | 2 | 3 | 1 | -- | 1 | -- | 5 | 1 |
| Thermal IR | -- | 1 | -- | -- | -- | -- | -- | 3 |
| Scene size (km) | 185 | 185 | 60 | 60 | 141-148 | 70 | 75 | 2700 |
| First Available | 1972 | 1982 | 1986 | 1986 | 1988 | 1995 | 1992 | ? |
| Relative Cost | $$ | $$$ | $$$$ | $$$$ | $$$ | $$$ | ? | $ |
| Contact in US | USGS | EOSAT | SPOT | SPOT | EOSAT | EOSAT | -- | USGS |
Landsat is the
oldest and most-used series of satellites for remote sensing of
Earth resources. Landsat is a series of satellites owned by
the United States. The first Landsat was launched in 1972,
and was originally called ERTS. A total of five Landsats
have orbited Earth. The first four have been retired, and
only Landsat 5 continues to operate today (as of 1997).
Landsat 6 was launched in 1996, but disappeared shortly after
takeoff, apparently becoming part of the oceans! Landsat 7
is under development and tentatively will be launched in
1998. Hopefully Landsat 5 will continue working until then,
though other satellites have recently been launched that gather
similar imagery.
Landsat satellites have carried two different sensors.
The first, Multispectral Scanner (MSS), has been on every
Landsat. MSS gathers low-resolution data compared to more
recent sensors: its pixels cover a ground area of about 80
meters (260 feet) on each side. It also detects energy in
only four bands: green, red, and two near-infrared bands.
Two good things about MSS is that data are available back to
1972, and they are relatively cheap for older scenes ($200, from
the
USGS Eros Data
Center in Sioux Falls, South Dakota).
Landsat Thematic Mapper (TM) has been on the satellites
since 1982. TM data are much more refined, since the pixels
cover 30 meters (100 feet) on each side. The sample image
here shows a portion of the Atlanta area with TM data. TM
also records seven bands of the spectrum, including blue and some
of the mid- and far-infrared. Unfortunately, TM data are
also more expensive. A private corporation,
Space Imaging (formerly EOSAT),
was formed by the US to market TM imagery. The charge for
scenes runs into the thousands of dollars.
In 1986, France launched
the first SPOT satellite (SPOT = Systeme Probatoire
d'Observation de la Terre). Three SPOT satellites have
been launched in total. They collect imagery similar to
Landsat TM, but the spatial resolution is finer. SPOT's
pixels are 20 meters for multispectral data. SPOT also
carries a panchromatic sensor, that is, the sensor records green
and red together in a single signal, similar to black-and-white
film. The panchromatic imagery has 10-meter pixels, half
the size of the multispectral SPOT sensor. The sample image
at right shows the same area as the TM image above, illustrating
the finer pixel resolution of SPOT panchromatic data. Until
very recently,
SPOT's imagery was
the best available to civilians. It's no surprise that it's
expensive -- over $5,000 per scene from what I understand!
(Gets into the territory of "If you have to ask, you can't
afford it.")
India and Japan have created their own satellites for gathering Earth data. The Indian program is ambitious, with three satellites since 1988. One sensor is almost identical to Landsat TM. Another offers the smallest pixels available today at just over 5 meters (8 feet). The Japanese satellite also offers high-quality imagery. Hopefully these new satellites will help make the cost of imagery more reasonable.
This sensor is unlike
those above in that it is much coarser in resolution. The Advanced
Very-High Resolution Radiometer (AVHRR) sensor is carried
aboard two US National Oceanic and Atmospheric Administration
(NOAA) weather satellites. The AVHRR pixels cover a ground
area that is one kilometer on each side -- 100 times larger than
SPOT panchromatic data. AVHRR is used for regional rather
than local studies, particularly of vegetation. The
illustration at right shows a composite of AVHRR images -- many
images have been pasted together to give an overall view of
Earth. The image below shows an AVHRR image around the Great
Lakes in an color-infrared style.
AVHRR data are very inexpensive. Several CD-ROMs are
easily available with a large amount of AVHRR data for about
$60. A few images are even available on the Internet
through the
USGS
Eros Data Center.
Several countries are planning new satellites similar to those above, most with smaller pixel sizes. Germany, Russia, China, Brazil, and Argentina have all announced satellite plans, in addition to the countries included above. NASA is planning several next-generation satellites that will carry hyperspectral sensors that will record up to 384 bands of the spectrum.
Private companies have also planned satellites for strictly commercial ventures. One, Space Imaging, intends to launch a satellite in the next few months that will gather data at a 1-meter pixel resolution. If this one works, it will rival aerial photography in detail. Until recently the US banned such detailed satellite imagery for defense reasons.
10. Assume you want to study wetlands areas and want to be able to see objects larger than 40 meters (130 feet) in size. You also want to be able to use both red and near infrared bands to detect vegetation levels. Which imagery product(s) among those discussed above might give you satisfactory data at a minimum cost?
Bryan Baker, Sonoma State
University, bryan.baker@sonoma.edu
Updated 17 February 1999