landsat 8 bands combination

17
Landsat 8 Bands Combination Summary Landsat 8 is the most recent satellite in the Landsat program. The data quality (signal-to-noise ratio) and radiometric quantization (12-bits) of the Landsat 8 Operational Land Imager (OLI) and Thermal Infrared Sensor (TIRS) are higher than previous Landsat instruments (8-bit for TM and ETM+). Since it's launch on February 11, 2013, Landsat 8 has been providing some truly stunning images of the earth's surface. Beyond their beauty, these images are packed with information which can be manipulated to extract features and discern changes to the earth's surface over time. When working with Landsat imagery, a logical first step is to load an image into an image analysis program and begin to visualize what is in the scene. The OLI sensor aboard Landsat 8 has nine bands for capturing the spectral response of the earth's surface at discrete wavelengths along the electromagnetic spectrum. Additionally, the TIRS sensor aboard Landsat 8 collects information at two discrete wavelengths within the thermal infrared portion of the electromagnetic spectrum. These wavelengths have been chosen carefully based on years of scientific research. Here’s how the new bands from Landsat 8 line up with Landsat 7: Landsat 7 Landsat 8 Band Name Bandwidt h (µm) Resolut ion (m) Band Name Bandwid th (µm) Resolut ion (m) Band 1 Coastal 0.43 – 0.45 30 Band 1 Blue 0.45 – 0.52 30 Band 2 Blue 0.45 – 0.51 30 Band 2 Green 0.52 – 0.60 30 Band 3 Green 0.53 – 0.59 30 Band 3 Red 0.63 – 0.69 30 Band 4 Red 0.64 – 0.67 30 Band 4 NIR 0.77 – 0.90 30 Band 5 NIR 0.85 – 0.88 30 Band 5 SWIR 1 1.55 – 1.75 30 Band 6 SWIR 1 1.57 – 1.65 30 Band 7 SWIR 2 2.09 – 2.35 30 Band 7 SWIR 2 2.11 – 2.29 30

Upload: nida-humaida

Post on 10-Dec-2015

522 views

Category:

Documents


31 download

DESCRIPTION

Summary

TRANSCRIPT

Page 1: Landsat 8 Bands Combination

Landsat 8 Bands Combination

Summary

Landsat 8 is the most recent satellite in the Landsat program. The data quality (signal-to-noise ratio) and radiometric quantization (12-bits) of the Landsat 8 Operational Land Imager (OLI) and Thermal Infrared Sensor (TIRS) are higher than previous Landsat instruments (8-bit for TM and ETM+). Since it's launch on February 11, 2013, Landsat 8 has been providing some truly stunning images of the earth's surface. Beyond their beauty, these images are packed with information which can be manipulated to extract features and discern changes to the earth's surface over time. When working with Landsat imagery, a logical first step is to load an image into an image analysis program and begin to visualize what is in the scene. The OLI sensor aboard Landsat 8 has nine bands for capturing the spectral response of the earth's surface at discrete wavelengths along the electromagnetic spectrum. Additionally, the TIRS sensor aboard Landsat 8 collects information at two discrete wavelengths within the thermal infrared portion of the electromagnetic spectrum. These wavelengths have been chosen carefully based on years of scientific research. Here’s how the new bands from Landsat 8 line up with Landsat 7:

Landsat 7 Landsat 8

Band Name

Bandwidth (µm)

Resolution (m)

Band Name

Bandwidth (µm)

Resolution (m)

Band 1 Coastal

0.43 – 0.45 30

Band 1 Blue

0.45 – 0.52 30

Band 2 Blue

0.45 – 0.51 30

Band 2 Green

0.52 – 0.60 30

Band 3 Green

0.53 – 0.59 30

Band 3 Red

0.63 – 0.69 30

Band 4 Red

0.64 – 0.67 30

Band 4 NIR

0.77 – 0.90 30

Band 5 NIR

0.85 – 0.88 30

Band 5 SWIR 1

1.55 – 1.75 30

Band 6 SWIR 1

1.57 – 1.65 30

Band 7 SWIR 2

2.09 – 2.35 30

Band 7 SWIR 2

2.11 – 2.29 30

Page 2: Landsat 8 Bands Combination

Band 8 Pan

0.52 – 0.90 15

Band 8 Pan

0.50 – 0.68 15

Band 9 Cirrus

1.36 – 1.38 30

Band 6 TIR

10.40 – 12.50 30/60

Band 10 TIRS 1

10.6 – 11.19 100

Band 11 TIRS 2

11.5 – 12.51 100

Landsat 8 measures different ranges of frequencies along the electromagnetic spectrum – a color, although not necessarily a color visible to the human eye. Each range is called a band, and Landsat 8 has 11 bands. Landsat numbers its red, green, and blue sensors as 4, 3, and 2, so when we combine them we get a true-color image such as this one:

Page 3: Landsat 8 Bands Combination

Landsat 8 view of the Los Angeles area, May 13th, 2013. The image is rotated so north is up. All image data courtesy of the U.S. Geological Survey. Have a look at the full list of Landsat 8’s bands:

Band Number µm Resolution

1 0.433–0.453 30 m

2 0.450–0.515 30 m

Page 4: Landsat 8 Bands Combination

3 0.525–0.600 30 m

4 0.630–0.680 30 m

5 0.845–0.885 30 m

6 1.560–1.660 30 m

7 2.100–2.300 30 m

8 0.500–0.680 15 m

9 1.360–1.390 30 m

10 10.6-11.2 100 m

11 11.5-12.5 100 m

Of its 11 bands, only those in the very shortest wavelengths (bands 1–4 and 8) sense visible light – all the others are in parts of the spectrum that we can’t see. The true-color view from Landsat is less than half of what it sees. To understand the value of all the bands, let’s look at them each in turn: The Bands Band 1 senses deep blues and violets. Blue light is hard to collect from space because it’s scattered easily by tiny bits of dust and water in the air, and even by air molecules themselves. This is one reason why very distant things (like mountains on the horizon) appear blueish, and why the sky is blue. Just as we see a lot of hazy blue when we look up at space on a sunny day, Landsat 8 sees the sky below it when it looks down at us through the same air. That part of the spectrum is hard to collect with enough sensitivity to be useful, and Band 1 is the only instrument of its kind producing open data at this resolution – one of many things that make this satellite special. It’s also called the coastal/aerosol band, after its two main uses: imaging shallow water, and tracking fine particles like dust and smoke. By itself, its output looks a lot like Band 2 (normal blue)’s, but if we contrast them and highlight areas with more deep blue, we can see differences:

Page 5: Landsat 8 Bands Combination

Band 1 minus Band 2. The ocean and living plants reflect more deep blue-violet hues. Most plants produce surface wax (for example, the frosty coating on fresh plums) as they grow, to reflect harmful ultraviolet light away. Bands 2, 3, and 4 are visible blue, green, and red. But while we’re revisiting them, let’s take a reference section of Los Angeles, with a range of different land uses, to compare against other bands:

Part of the western LA area, from agricultural land near Oxnard in the west to Hollywood and downtown in the east. Like most urban areas, the colors of the city average out to light gray at this scale. Band 5 measures the near infrared, or NIR. This part of the spectrum is especially important for ecology because healthy plants reflect it – the water in their leaves scatters the wavelengths back

Page 6: Landsat 8 Bands Combination

into the sky. By comparing it with other bands, we get indexes like NDVI, which let us measure plant health more precisely than if we only looked at visible greenness.

The bright features are parks and other heavily irrigrated vegetation. The point near the bottom of this view on the west is Malibu, so it’s a safe bet that the little bright spot in the hills near it is a golf course. On the west edge is the dark scar of a large fire, which was only a slight discoloration in the true-color image. Bands 6 and 7 cover different slices of the shortwave infrared, or SWIR. They are particularly useful for telling wet earth from dry earth, and for geology: rocks and soils that look similar in other bands often have strong contrasts in SWIR. Let’s make a false-color image by using SWIR as red, NIR as green, and deep blue as blue (technically, a 7-5-1 image):

The fire scar is now impossible to miss – reflecting strongly in Band 7 and hardly at all in the others, making it red. Previously subtle details of vegetation also become clear. It seems that plants in the canyons north of Malibu are more lush than those on the ridges, which is typical of climates where water is the main constraint on growth. We also see vegetation patterns within LA – some neighborhoods have more foliage (parks, sidewalk trees, lawns) than others.

Page 7: Landsat 8 Bands Combination

Band 8 is the panchromatic – or just pan – band. It works just like black and white film: instead of collecting visibile colors separately, it combines them into one channel. Because this sensor can see more light at once, it’s the sharpest of all the bands, with a resolution of 15 meters (50 feet). Let’s zoom in on Malibu at 1:1 scale in the pan band:

And in true color, stretched to cover the same area:

Page 8: Landsat 8 Bands Combination

The color version looks out of focus because those sensors can’t see details of this size. But if we combine the color information that they provide with the detail from the pan band – a process called pan sharpening – we get something that’s both colorful and crisp:

Pansharpened Malibu, 15 m (50 ft) per pixel. Notice the wave texture in the water. Band 9 shows the least, yet it’s one of the most interesting features of Landsat 8. It covers a very thin slice of wavelengths: only 1370 ± 10 nanometers. Few space-based instruments collect this part of the spectrum, because the atmosphere absorbs almost all of it. Landsat 8 turns this into an advantage. Precisely because the ground is barely visible in this band, anything that appears clearly in it must be reflecting very brightly and/or be above most of the atmosphere. Here’s Band 9 for this scene:

Page 9: Landsat 8 Bands Combination

Band 9 is just for clouds! Here it’s picking up fluffy cumulus clouds, but it’s designed especially for cirrus clouds – high, wispy “horsetails”. Cirrus are a real headache for satellite imaging because their soft edges make them hard to spot, and an image taken through them can contain measurements that are off by a few percent without any obvious explanation. Band 9 makes them easy to account for. Bands 10 and 11 are in the thermal infrared, or TIR – they see heat. Instead of measuring the temperature of the air, like weather stations do, they report on the ground itself, which is often much hotter.

Important TIRS calibration notice from USGS, Jan. 6, 2014: “Due to the larger calibration uncertainty associated with TIRS band 11, it is recommended that users refrain from relying on band 11 data in quantitative analysis of the TIRS data, such as the use of split window techniques for atmospheric correction and retrieval of surface temperature values.” We suggest that Band 10 be used in conjunction with an atmospheric model to estimate surface brightness temperature. Our calibration team has found that with current processing these surface brightness temperatures are accurate to within ~±1 K for many 15 – 35° C targets, e.g., growing season vegetated targets. For more details, please visit the Landsat 8 Data Users Handbook

A study a few years ago found some desert surface temperatures higher than 70 °C (159 °F) – hot enough to fry an egg. Luckily, LA is relatively temperate in this scene:

Page 10: Landsat 8 Bands Combination

Notice that the very dark (cold) spots match the clouds in Band 9. After them, irrigated vegetation is coolest, followed by open water and natural vegetation. The burn scar near Malibu, which is covered in charcoal and dry, dead foliage, has a very high surface temperature. Within the city, parks are generally coolest and industrial neighborhoods are warmest. There is no clear urban heat island in this scene – an effect that these TIR bands will be particularly useful for studying. Bands can be combined in many different ways to reveal different features in the landscape. Let’s make another false-color image by using this TIR band for red, a SWIR band for green, and the natural green band for blue (a 10-7-3 image):

Page 11: Landsat 8 Bands Combination

Urban areas and some kinds of soil are pink. In the true-color image, wild vegetation is almost uniformly olive-colored, but here we see a distinction between peach-colored scrubland, mahogany-colored woodland, and so on. Cooling onshore breezes appear as a slight purple gradient along the coast of the city. The colored strips on either side of the image are areas where not all sensors have coverage. Standard digital cameras are designed to replicate what we see with the human eye, so they capture light only in the red, green and blue wavelengths and then apply red, green and blue filters (also known as channels) to these wavelengths, respectively, that when combined generate a natural looking RGB image. With a multispectral image from a sensor system such as Landsat 8, we have a lot more information to work with. Different wavelengths can often help us discern some features better than others or even help us "see through" features such as clouds or smoke. For example, the Near Infrared (NIR) wavelength is one of the most commonly used wavelengths on multispectral sensors because vegetation reflects so strongly in this portion of the electromagnetic spectrum that this information proves very useful when performing vegetation analyses. The Shortwave Infrared (SWIR) bands aboard Landsat 8 are very useful for discerning differences in bare earth and for telling what is wet and what is dry in a scene. There are many other examples of the advantages of the available bands in Landsat images, but what I would like to do here is simply show how loading different combinations of these bands into the red, green and blue channels makes different features stand out. I am not the first to do this, but I just thought I would add an additional resource to the world wide web for showing how these band combinations can be used to visualize Landsat 8 images.

Natural Color 4 3 2

Page 12: Landsat 8 Bands Combination

False Color (urban) 7 6 4

Color Infrared (vegetation) 5 4 3

Agriculture 6 5 2

Atmospheric Penetration 7 6 5

Healthy Vegetation 5 6 2

Land/Water 5 6 4

Natural With Atmospheric Removal 7 5 3

Shortwave Infrared 7 5 4

Vegetation Analysis 6 5 4

4 , 3 , 2 - Natural Color Image, Fresno, California This band combination is as close to "true color" as you can get with a Landsat OLI image. One unfortunate drawback with this band combination is that these bands tend to be susceptible to atmospheric interference, so they sometimes appear hazy.

5, 4, 3 - Traditional Color Infrared (CIR) Image, Colorado/Utah Note how vegetation really pops in red, with healthier vegetation being more vibrant. It's also easier to tell different types of vegetation apart than it is with a natural color image. This is a very commonly used band combination in remote sensing when looking at vegetation, crops and wetlands.

Page 13: Landsat 8 Bands Combination

7, 6, 4 - False Color useful for visualizing urban environments, Los Angeles, California Because this band combination makes use of both of the SWIR bands aboard Landsat 8, the image is much more crisp than band combinations that make use of bands in shorter wavelengths, which are more susceptible to haze.

5, 6, 4 - False Color good for picking out land from water, Hudson Bay, Canada In this false color image, land appears in shades of orange and green, ice stands out as a vibrant magenta color, and water appears in shades of blue.

Page 14: Landsat 8 Bands Combination

7, 5, 3 - False color image with good atmospheric penetration, Washington/Oregon This band combination is similar to the 5, 6, 4 band combination shown above, but vegetation shows up in more vibrant shades of green. This band combination was used for the global Landsat mosaic created by NASA.

6, 5, 2 - False color for agriculture, Fruita, Colorado This band combination is useful for the monitoring of agricultural crops, which appear as a vibrant green. Bare earth appears as a magenta color and non-crop vegetation appears as more subdued shades of green.

Page 15: Landsat 8 Bands Combination

7, 5, 2 - False color often used for visualizing forest fire burn scars, Rim Fire, California This band combination is similar to the 6, 5, 2 band combination shown above, but by pushing further into the SWIR range of the electromagnetic spectrum, there is less susceptibility to smoke and haze generated by a burning fire.

6, 3, 2 - False color for distinguishing differences in bare earth, Canyonlands NP, Utah This band combination is good for discerning variations in a landscape that does not contain an abundance of vegetation. It is good for geologic applications.

Page 16: Landsat 8 Bands Combination

5, 7, 1 - False color for vegetation and water, Lake Victoria, Tanzania This band combination makes use of the NIR, SWIR2, and Coastal Aerosol bands, respectively. The Coastal Aerosol band is unique to Landsat 8 and is used primarily to track fine particles like dust and smoke, and also to peer into shallow water. With this color combination, vegetation appears orange.

There are certainly several additional band combinations that would be useful for visualizing Landsat 8 scenes. In some situations, loading a grayscale image of a single band might also help to visualize specific features or phenomena. For example, in the grayscale image below, we are viewing the Thermal Infrared 1 band from the Landsat 8 TIRS sensor. This image, captured over Iceland, shows the Bardarbunga volcano in bright white in the top center of the image. Since we are using a thermal infrared band, this feature sticks out significantly from the ice sheet located just to the south of the lava flow.

Page 17: Landsat 8 Bands Combination

With grayscale images, we also have the option of applying a color table to highlight features a little more clearly. In the image below we are viewing the same grayscale image of the Bardarbunga volcano, but now we have applied a color table that shows very cold extremes of the glacier in shades of deeper blue and very warm extremes of the lava flow in shades of deeper red.

There are many analysis techniques that can be applied to multispectral imagery to extract specific features of interest. These algorithms rely on the same principles of reflectivity and absorption at various wavelengths that allow us to see certain features when visualizing them with different band combinations. An important point to note is that if a particular band, or combination of bands, does a good job of helping you visualize a feature that is of interest to you, then it is highly likely that this band, or combination of bands, can be used to help you isolate that feature from your image. For more on this topic, please check out the section of our Documentation Center devoted to Spectral Indices.