See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/311901147

Exploring Landsat 8

Article · May 2015

CITATIONS READS

10 6,513

2 authors:

Tri Dev Acharya Intae Yang

University of California, Davis Kangwon National University
58 PUBLICATIONS   383 CITATIONS    56 PUBLICATIONS   220 CITATIONS   

SEE PROFILE SEE PROFILE

Some of the authors of this publication are also working on these related projects:

Geographic Information Systems for Transportation (GIS-T) in Nepal View project

Mentor4Nepal Initiative View project

All content following this page was uploaded by Tri Dev Acharya on 09 February 2017.

The user has requested enhancement of the downloaded file.

International Journal of IT, Engineering and Applied Sciences Research (IJIEASR) ISSN: 2319-4413 4
Volume 4, No. 4, April 2015

Exploring Landsat 8
Tri Dev Acharya, Graduate Student, Department of Civil Engineering, Kangwon National University, Chuncheon,
South Korea
Intae Yang, Professor, Department of Civil Engineering, Kangwon National University, Chuncheon, South Korea

ABSTRACT possibly with frequent cloud cover or for a certain time of

the year often good images cannot be found.
Landsat has been providing moderate spatial resolution,
global, synoptic, and repetitive coverage of the Earth’s With the launch of Landsat 8 in February of this year, the
land surfaces. Landsat 8 extends the remarkable 40 year continuity of the program is assured into at least the next
Landsat record and has enhanced capabilities. This paper decade. And in case of successful operation, the
introduces overview of sensors and their standard data refinements and new bands will overcome the previous
products Preliminary evaluation of Landsat 8 capabilities necessities with more detail and accuracy (USGS, 2013).
and identification of new science and applications
opportunities are explored. 2. LANDSAT 8 PROGRAM

Keywords The Landsat Data Continuity Mission has a successful

Landsat, Landsat 8, Satellite Imagery, Band Combination. launch February 11th 2013. It was officially renamed to
Landsat 8 on May 30, 2013. The newest satellite in the
Landsat series offers scientists a clearer view with better
1. INTRODUCTION spatial resolution than most ocean-sensing instruments and
greater sensitivity to brightness and color than previous
Satellite observations of Earth’s land and ocean surfaces Landsats. Most significantly, it can observe the Earth in
are very important for understanding the processes and wavelengths that allows scientists to adjust for the
rhythms of our planet. There has been many such satellites distortions especially caused by the atmosphere near the
has been providing services every hour. Many new such coast (http://earthobservatory.nasa.gov/IOTD/).
advanced programs are being added year by year. Among
the most successful and reliable program has been Landsat
program.

The Landsat era began in 1972, providing moderate spatial

resolution, global, synoptic, and repetitive coverage of the
Earth’s land surfaces, continues at a scale where natural
and human-induced changes can be detected,
differentiated, characterized, and monitored over time
(USGS, 2013). Landsat data from the United States
Geological Survey (USGS) is one of the best sources for
mapping and monitoring of land cover and land surface Fig. 1. Bandpass wavelengths for the landsat 8 OLI
biophysical and geophysical properties over the last 40 and TIRIS sensors, compared to Landsat 7 ETM+
years (Hansen & Loveland, 2012; Wulder et. al., 2012). Of sensors (USGS, 2013)
the last three Landsats, Landsat 7 reached orbit but in 2003 Landsat 8 carries two instruments: The Operational Land
incurred a problem that created ongoing striping of Imager (OLI) and Thermal Infrared Sensor (TIRS). The
missing data on imagery, Landsat 6 did not reach orbit, OLI, a push-broom sensor with a four mirror telescope,
and Landsat 5 operated successfully well beyond its collects data in visible, near infrared, and shortwave
expectancy. infrared wavelength regions as well as a panchromatic
band. Two new spectral bands have been added: a deep-
Landsat 5 which produced images since the 1980s but at a blue band for coastal water and aerosol studies (band 1),
fairly low rate and which stopped delivering image with its and a band for cirrus cloud detection (band 9) (Fig. 1). A
main sensor in 2011. Landsat 7 before the SLC failure in Quality Assurance band is also included to indicate the
2003 produced a good and high quality coverage of most presence of terrain shadowing, data artifacts, and clouds
of the earth but these are meanwhile more than 10 years (USGS, 2013). Panchromatic, multispectral are taken in
old so can be outdated for a lot of applications. As the 15m, 30m resolution respectively. The TIRS collects data
coverage is limited, images from more remote regions are in two long wavelength thermal infrared bands. The 100-
meter spatial resolution of TIRS data is registered to the

i-Explore International Research Journal Consortium www.irjcjournals.org

International Journal of IT, Engineering and Applied Sciences Research (IJIEASR) ISSN: 2319-4413 5
Volume 4, No. 4, April 2015

OLI data to create radiometrically and geometrically enable better characterization of land cover state and
calibrated, terrain-corrected 16-bit Level 1 data products. condition.
Data collected by the instruments onboard the satellite, are
Table 1. Processing parameter for Landsat 8 standard available to download at no charge from GloVis,
data products (USGS, 2013) EarthExplorer, or via the LandsatLook Viewer within 24
Product Type Level 1T (terrain corrected) hours of reception (USGS, 2013). Approximate scene size
Data type 16-bit unsigned integer is 170 km north-south by 183 km east-west (106 mi by
Output format GeoTIFF 114 mi) and have a large file size of about 1 GB
Pixel size 15 meters/30 meters/100 meters compressed.
(panchromatic/multispectral/thermal)
Map UTM (Polar Stereographic for Landsat 8 data products are consistent with the all standard
projection Antarctica) Level-1 (orthorectified) data products created using
Datum WGS 84 Landsat 1 to Landsat 7 data. It consists of quantized and
Orientation North-up (map) calibrated scaled Digital Numbers (DN) representing
Resampling Cubic convolution multispectral image data acquired by both the Operational
Accuracy OLI: 12 meters circular error, 90 % Land Imager (OLI) and Thermal Infrared Sensor (TIRS).
confidence The products are delivered as 16-bit unsigned integer
format (scaled to 55,000 grey levels), and can be rescaled
TIRS: 41 meters circular error, 90 %
to the Top Of Atmosphere (TOA) reflectance and/or
confidence
radiance using radiometric rescaling coefficients provided
in the product metadata file (MTL file). The MTL file also
On March 18, 2013, the first images were taken with both
contains the thermal constants needed to convert TIRS
the Operational Land Imager (OLI) and the Thermal
data to the at-satellite brightness temperature.
Infrared Sensor (TIRS) onboard the Landsat Data
Continuity Mission (LDCM) spacecraft. The Landsat 8
The level of detail (spatial resolution) is often the most
satellite images the entire Earth every 16 days in an 8-day
interesting aspect of viewing a satellite image, but less
offset from Landsat 7. It promises to produce images at a
appreciated is how changes in irradiative energy reflected
significantly higher rate than Landsat 7 (up to 400 scenes
by different surface materials are used to identify features
per day). Both sensors provide improved signal-to-noise
of interest. Landsat 8 provides continuity to the previous
(SNR) radiometric performance quantized over a 12-bit
sensors and also with its addition of new multispectral and
dynamic range. Improved signal to noise performance
thermal bands enables additional analysis in future.
Table 2. Landsat 8 bands properties and applications
Landsat 8 Band Band Name Wavelength Resolution Applications
Sensors (µm) (m)
1 Coastal / Aerosol 0.433-0.453 Coastal and Aerosol studies
Bathymetric mapping, distinguishing soil from
2 Blue vegetation and deciduous from coniferous
0.450-0.515 vegetation
Emphasizes peak vegetation, which is useful for
3 Green
0.525-0.600 assessing plant vigour
Operational
4 Red 0.630-0.680 30 Discriminates vegetation slopes
Land
Near Infrared
Imager 5 Emphasizes biomass content and shorelines
(NIR) 0845.-0.885
(OLI)
Short-wave Discriminates moisture content of soil and
6
Infrared (SWIR) 1 1.560-1.660 vegetation; penetrates thin clouds
Short-wave Improved moisture content of soil and vegetation
7
Infrared (SWIR) 2 2.100-2.300 and thin cloud penetration
8 Panchromatic 0.500-0.680 15 Sharper image definition
9 Cirrus 1.360 -1.390 30 Improved detection of cirrus cloud contamination
Thermal Long-wave 10.30 –
10 Thermal mapping and estimated soil moisture
Infrared Infrared (LWIR) 1 11.30
100*
Sensor Long-wave Improved thermal mapping and estimated soil
11
(TIRS) Infrared (LWIR) 2 11.50-12.50 moisture
Quality
BQA Quality assessments for every pixel in the scene
Assessment
* TIRS bands are acquired at 100 meter resolution, but are resampled to 30 meter in delivered data product.

i-Explore International Research Journal Consortium www.irjcjournals.org

International Journal of IT, Engineering and Applied Sciences Research (IJIEASR) ISSN: 2319-4413 6
Volume 4, No. 4, April 2015

3. PROPERTIES necessary to apply the atmospheric water absorption

correction factor.
In Landsat 8, the heritage bands are similar but refined to
avoid atmospheric absorption features (enabled by higher The Coastal/Aerosol is blue water penetrating and aerosol
signal to noise ratio inherent in push-broom architecture). detecting band. It has been added at request of ocean color
Due to improvement, it is capable of providing a 12-bit investigators requiring higher resolution of coastal water
range (4,096 levels) of pixel values, improving precision. relative to MODIS and SeaWiFS. Relative to the MODIS
and the SeaWiFS sensor, the Landsat 8 sensor allows for
Table 3. Comparing the differences to previous better imaging of these shallow waters due to its superior
version of Landsat spatial and radiometric resolutions. By comparing with
Band Name Band Number Differences in blue band, correction for the difference can be applied
LS8 LS7 LS5 LS8 which will help in better and closer inspection of coastal
Coastal/ Aerosol 1 - - new and inland waters. It can also help in estimation of carbon
Blue 2 1 1 more narrow exchange at the land-water interface such as salt marshes,
green 3 2 2 more narrow wetlands, harbors where ocean color instruments fail due
to coarse pixel size. Scientists also hope to make clearer
red 4 3 3 more narrow
observations of sediments, particles, organic matter, coral
Near Infrared 5 4 4 more narrow
reefs, and suspended chlorophyll-rich phytoplankton in
(NIR)
these bodies of water. The same band, Coastal/Aerosol (its
Short-wave 6 5 5 more narrow
radiance value) can be used with two other bands for
Infrared (SWIR) 1
estimating the concentration of aerosols in the atmosphere,
Short-wave 7 7 7 more narrow which may be used to refine the atmospheric correction
Infrared (SWIR) 2 procedures such as dark object subtraction.
Panchromatic 8 8 - more narrow,
only visible Another new band, Cirrus allows for better detection of
(red-green) cirrus cloud contamination in each scene. The old Landsat
Cirrus 9 - - new sensors included bands that made it hard to detect these
Long-wave 10 6 6 two bands high-altitudes, cold, and wispy difficult-to-detect clouds.
Infrared (LWIR) 1 instead of one This new cirrus band with center wavelength of 1.375
Long-wave 11 6 6 two bands microns attempts to detect the light, which is reflected by
Infrared (LWIR) 2 instead of one the high-altitude clouds but is absorbed by the water vapor
closer to the ground. In this band, cirrus clouds will appear
The thermal infrared band from Landsat 7 is now split into bright while most land surfaces will appear dark through
two bands for Landsat 8. Whereas before you had one the atmospheres that contain water vapor. If the
thermal band that was acquired at 60 m resolution (and atmosphere is relatively dry, the one risk having some of
resampled to 30 m) now you have increased spectral the land surfaces appear bright as well. Using the cirrus
resolution at the cost of spatial resolution. Both bands 10 band requires careful interpretation techniques. This band
and 11 are useful in providing more accurate surface is not to be used together with other multispectral bands
temperatures and support emerging applications such as but, as a cloud masking.
modeling evapotranspiration for monitoring water use
consumption over irrigated lands. Similarly, the new Quality Assessment band represents
bit-packed combinations of surface, atmosphere, and
The new infrared bandwidth is more improved and more sensor conditions that can affect the overall usefulness of a
uniform across the scene. It will not get absorbed as much given pixel. It provides weather a pixel is good or bad to
by atmospheric water as NIR bands of previous Landsats. produce more accurate and precise results for each user’s
As shown in the Method, guide, and accuracy page in the application. It also helps in differentiating between snow
Atmospheric Correction folder, an adjustment is needed to covered mountains and highly reflective urban areas
correct the NIR band for Landsat 5 and 7 due to absorption (USGS, 2014).
of the radiation by atmospheric perceptible water (Wu et Some of Landsat 8’s bands exhibit striping in the images,
al., 2005) to retrieve accurate enough reflectance. The probably due to sensor calibration problems. So far the
refinement of the Landsat 8 NIR bandwidth helps solve thermal, cirrus, and coastal aerosol bands have been
this problem. Landsat 7 NIR (band 4) bandwidth is identified as those that show striping most prominently.
0.775–0.900 µm while Landsat 8 NIR (band 5) bandwidth The stripes in the thermal bands are the most problematic,
0.845–0.885 µm. Landsat 8 NIR is in much more of an because even though they are narrow and short in length,
atmospheric window in regards to absorption by water; they show the biggest offset in pixel values. The stripes in
radiation will not get absorbed enough for it to be the cirrus and coastal aerosol bands are easier to detect
since they are broad and long. However, the offsets in

i-Explore International Research Journal Consortium www.irjcjournals.org

International Journal of IT, Engineering and Applied Sciences Research (IJIEASR) ISSN: 2319-4413 7
Volume 4, No. 4, April 2015

pixel values in these two bands are minimal (Yale Imagery data for the study were acquired from the United
University, 2013). States Geological Survey (USGS) website using its online
interface, the GLOVIS (http://glovis.usgs.gov/). The data
4. APPLICATION were acquired as zipped Georeferenced Tagged Image File
Format (GeoTIFF) representing systematically terrain
A. Study Scene corrected data (L1T). The properties and the natural color
The study scene is North East part of South Korea. It image and quality assessment band of scene are as shown
consists mostly of Gangwon province and some part of in table 4 and Fig. 2.
Gyeonggi, Chungcheongbuk and Gyeongsangbuk
provinces. It also includes DMZ boarder and small part of The software used for processing and information
North Korea. extraction are ArcGIS 10.2 (ESRI, 2013) and ENVI 5.1
(EVIS, 2013). Both of the software support Landsat
B. Data set and methodology metadata file.

Fig. 2. Scene from path 34 and row 115 a. Natural color composite (432); b. Quality Assessment Band

Table 4. Details of the study scene X

Data Set Attribute Attribute Value Geometric RMSE Model 6.036
Landsat Scene Identifier LC81150342013300LGN00 Y
WRS Path 115 Browse Exists Yes
WRS Row 034 Center 37°28'27.88"N,
Nadir Off Nadir NADIR 128°11'42.65"E
Full or Partial Scene FULL NW Corner 38°31'42.06"N,
Data Category NOMINAL 127°23'25.98"E
Roll Angle -0.001 NE Corner 38°07'46.92"N,
Station Identifier LGN 129°30'41.51"E
Day/Night DAY SE Corner 36°24'21.35"N,
Data Type Level 1 Level 1T 128°58'21.79"E
Sensor Identifier OLI_TIRS SW Corner 36°48'11.84"N,
Date Acquired 10/27/2013 126°53'59.93"E
Start Time 2013:300:02:06:20.0758020
Stop Time 2013:300:02:06:52.0761252
Image Quality 9 5. RESULTS
Scene Cloud Cover 1.31
Sun Elevation 37.67814162 The GeoTIF imageries were opened with the metadata file
Sun Azimuth 159.9862116 in ENVI 5.1. It categorized all separate bands into
Geometric RMSE Model 5.013 Multispectral (1, 2, 3, 4, 5 6 and 7), Panchromatic (8),
Cirrus (9), Thermal Infrared (10, 11) and Quality (BQA).

i-Explore International Research Journal Consortium www.irjcjournals.org

International Journal of IT, Engineering and Applied Sciences Research (IJIEASR) ISSN: 2319-4413 8
Volume 4, No. 4, April 2015

Each group was exported to individual set. Radiometric

calibration was performed to convert DN value of
multispectral imagery to At-sensor reflectance. After
calibration, bands were explored individually or in
different combinations.

Fig 5. a. Band 6; b. Image comparison (diagonal) with

original 432 resolutions and pansharpened image

Panchromatic band is black and white band which collects

energy from visible spectrum combined. As combined
helps in more collection of light, the images are sharper,
Fig 3. a. Band 1; b. Difference between Band 1 and which is actually 15m resolution. These are used in
Band 2 pansharpening the regular lower resolution images. Fig. 5
shows images with and without sharpening; the sharpened
Coastal/Aerosol band senses blue and violet. It is very image shows more crisp details than blur original
hard to collect as it gets scattered by small dust, water resolutions.
particle in air and even air molecules themselves. The
difference between band 1 and 2 shows the ocean and Cirrus band is one of the new and important features of
living plants reflect greener color. The effect in vegetation Landsat 8. It collects a very narrow wavelength: only 1370
is due to Epicuticular wax on growing plants to reflect ± 10 nanometers. Few space-based instruments collect this
harmful radiations. It also reveals some stripping effect part of the spectrum, because the atmosphere absorbs
(Fig.3). almost all of it. The ground is barely visible in this band
and anything that appears clearly in, it must be reflecting
The heritage bands Blue, Red and Green produce more very brightly and/or be above most of the atmosphere.
clearly natural color image. Fig. 4 shows the natural color
composite of Gangneung area. Sea looks dark blue Thermal infrared or TIR bands measures the temperature
whereas forests are green and brown. Urban areas are light on the ground itself. Analysing these bands, heat islands
grey. can be identified in urban areas. Both these bands show
stripping effect but band 11 shows more instability in
values than band 10. Fig 6.b shows the corner parts ocean
and city are little hotter than the mid forest hills. The river
channel shows moderate temperature.

Fig 4. Gangenung area a. Band 432; b. Band 5

Near Infrared band is small part of spectrum which is

refraction towards the sky form water content of healthy Fig 6. a. Band 9; b. Thermal band 10
plant leaves. Fig 4.b shows the bright part as densely As all bands are collected exactly the same time and place,
growing vegetation. Shortwave Infrared I and II are the all eleven bands can be used together. Allowing artificial
useful in determining the wetland from dry, differentiating colors to data form different spectral bands for display
different rock and soil types that look similar. helps scientists to distinguish between different surface
features according to their reflection intensity. Some
common band combinations applied to Landsat 8,
displayed as a red, green, blue (RGB) are shows in Fig. 7
and table 5.

i-Explore International Research Journal Consortium www.irjcjournals.org

International Journal of IT, Engineering and Applied Sciences Research (IJIEASR) ISSN: 2319-4413 9
Volume 4, No. 4, April 2015

Table 7. Application of various band combinations

Application Band Sample
Combinati Scene
on
False Colour (urban): The violet blue shows the city expansion. 764 Fig 7a
Colour Infrared (vegetation): One of the most used false colour image to identify 543 Fig 7b
vegetation cover.
Agriculture: The composite differentiate agricultural area in light yellow display. 652 Fig 7c
Atmospheric Penetration: The all infrared composite shows how much penetration is 765 Fig 7d
made through the atmosphere by the spectrum.
Healthy Vegetation: The composite helps in more detail status of vegetation health. 562 Fig 7e
Land/Water: Water bodies are easily separated with this composite. 564 Fig 7f
Natural With Atmospheric Removal: The composite gives the natural like view with 753 Fig 7g
the help of infrared atmospheric window.
Vegetation Analysis: Vegetation can be categories more easily and health conditions 654 Fig 7h
can also be analysed.
Thermal Variation: Water surfaces are redder i.e. hotter than land surface. Sea breeze 10 7 3 Fig 7i
caused the east shores hotter which is within mountain range of Taebaek Some light
blue cold areas are due to clouds. The coloured strips on either side of the image are
areas where not all sensors have coverage.

i-Explore International Research Journal Consortium www.irjcjournals.org

 International Journal of IT, Engineering and Applied Sciences Research (IJIEASR) ISSN: 2319-4413 10
Volume 4, No. 4, April 2015

Fig. 7. Band compositions: (a) 7 6 4; (b) 5 4 3; (c) 6 5 2; (d) 7 6 5; (e) 5 6 2; (f) 5 6 4; (g) 7 5 3; (h) 6 5 4; (i) 10 7 3

Beside band combination, ratios and indices are also very [4] USGS (2013), United States Geological Survey,
important for studying the earth surface, which could Landsat 8: U.S. Geological Survey Fact Sheet
reveal more detail and accurate detection for scientific 2013–3060, http://pubs.usgs.gov/fs/2013/3060/ (last
studies. date accessed: 25 November, 2014)
[5] USGS (2014), United States Geological Survey,
6. CONCLUSION Landsat 8 Quality Assessment Band,
https://landsat.usgs.gov/L8QualityAssessmentBand.
Landsat program has been providing space-based php (last date accessed: 28 November, 2014)
moderate-resolution land remote sensing data continuously [6] Wu, J., Wang, D., & Bauer, M. E. (2005). Image-
since four decades. The acquired imageries are the used in based atmospheric correction of QuickBird imagery
many different applications in the field of agriculture, of Minnesota cropland. Remote Sensing of
forestry, geology, regional planning, education, mapping, Environment, 99(3), 315-325.
and global landuse/cover change. With the launch of [7] Wulder, M. A., Masek, J. G., Cohen, W. B.,
Landsat 8, the legacy has been forwarded to future. Loveland, T. R., & Woodcock, C. E. (2012).
Opening the archive: How free data has enabled the
Landsat 8 has improved its features as well as introduced science and monitoring promise of Landsat. Remote
new bands for more accurate studies in broader fields. It Sensing of Environment, 122, 2-10.
offers more combinations with more narrow classification. [8] Yale University (2013), Yale Guide to Landsat 8
Yet, more has to be explores about Landsat 8, scientific Image Processing,
research as well as whole mankind will be benefited with http://www.yale.edu/ceo/Documentation/Landsat%
its improved imagery archives. 208%20image%20processing.pdf (last date
accessed: 2 December, 2014)

REFERENCES

[1] ESRI (2013), Environmental Systems Research

Institute, ArcGIS 10.2 for (Desktop, Engine,
Server) Landsat 8 Patch,
http://support.esri.com/en/downloads/patches-
servicepacks/view/productid/66/metaid/2012 (last
date accessed: 5 December, 2014)
[2] EVIS (2014), Exelis Visual Information Solutions,
Radiometric Calibration (Using ENVI),
http://www.exelisvis.com/docs/RadiometricCalibrat
ion.html (last date accessed: 2 December, 2014)
[3] Hansen, M. C., & Loveland, T. R. (2012). A review
of large area monitoring of land cover change using
Landsat data. Remote sensing of Environment, 122,
66-74.

i-Explore International Research Journal Consortium www.irjcjournals.org

View publication stats