Spectral Signature Cheatsheet – Spectral Bands in Remote ...

1 Spectral Signature Cheatsheet Spectral Bands in Remote Sensing Which Spectral Bands to use in Remote sensing? Spectral Reflectance in Remote Sensing Last Updated: Feb 20, 2018 @ 6:21 pm Explore the Visible (and Invisible) with Spectral Bands Last month, I started building a secret list, in my top secret laboratory, of Spectral Bands and their Spectral signatures. I asked the world s wisest to contribute their wisdom. The result? One of the best Spectral Signature cheatsheets that exists today. Today, we explore the visible and invisible. Whether its near-infrared or the cirrus band , use this clear and concise guide to Remote sensing Spectral signatures to extract as much data as possible from your imagery. What is a Spectral band , Anyhow? Thanks to our atmosphere, we only see specific portions of the electromagnetic (EM) spectrum. The graph below shows our atmospheric window . The EM energy (in blue) is what we are capable of seeing on Earth.

2 Our eyes can only see the visible portion red, green and blue. Healthy vegetation (or chlorophyll) reflects more green light compared to other wavelengths. It absorbs more red and blue light. That is why our eyes see it as green. But special types of sensors can pick up other forms of the EM spectrum invisible to the human eye. For example, vegetation also reflects even more near-infrared (NIR). Near-infrared is invisible to the human eye, but sensors can pick up this Spectral band ..So special types of sensors can reveal new insights about Earth s features, that our eyes can t. AVHRR NDVI Composite Image Courtesy of NOAA. About that Top Secret Getting back to the whole Spectral Signature top-secret My confession is that it s actually the USGS Spectral Signature laboratory. We don t have a Spectral band laboratory at ..But the USGS has and a handy Spectral Characteristics Viewer. And they ve vigorously tested the Spectral reflectance of hundreds of materials in the lab.

3 All this data has been compiled in their Spectral library website. The takeaway is this: Each and every object has its own chemical composition. This is the equivalent to saying that each composition has its own Spectral Signature . Unsupervised Classification Example Take the mining industry, there are over 4000 natural minerals on Earth. Each one has its own chemical composition that makes it different from one another. When you have more Spectral Bands , this gives greater potential to map more minerals, vegetation or anything! OK, what s your point? You can pick and choose which Spectral Bands to use in your image classification techniques in Remote sensing. This means in your Remote sensing classification you will have a higher probability to auto-magically pull out features of interest with Spectral signatures..which brings us to our next section the Spectral Signature Cheatsheet . The Spectral Signature Cheatsheet A Spectral Signature plots all the variations of reflected EM radiation as a function of wavelengths.

4 Pick and choose the Spectral Bands you use in your classification. If you want to pull out vegetation from a multispectral image, this Spectral Signature clearly shows you should leverage the near-infrared band ( m). Vegetation and water Spectral Signature line graph ..and Normalized Difference Vegetation Index (NDVI) might be just one of the best ways to normalize your data. Nothing is written in stone for image classification. These are just general guidelines to help you for classification. Without further 1 Coastal Aerosol ( m) As the name implies, your coastal aerosol band is particularly useful in coastal, bathymetric and aerosol studies. The coastal band reflects blues and violets and is becoming more common on satellite sensors such as Landsat-8 and WorldView-3. Here are some of its Spectral band applications: WHALE POPULATION: Because of its ability to penetrate water (up to 20-30m in clear water), the coastal band is being used to estimate (sub-surface) whale populations This Spectral band allows more confidence to the identification of whales in the wild as opposed to observations of animals in captive tanks.

5 (Counting whale populations from space) BATHYMETRY: Because the coastal aerosol band is absorbed less by water, it allows for coastal water and ocean color observation. This is also useful for Spectral signatures for marine vegetation like seagrass and other underwater benthic habitat. AEROSOLS: Because the coastal aerosol band is more sensitive to clouds, smoke and haze, it s being used for filtering out clouds in imagery processing. Earth observers are fine-tuning cloudless orthoimagery as basemaps with the help of the coastal aerosol band . (Cloud detection and removal techniques in Landsat 8 imagery) 2 Blue ( m) Snorkel in an oasis of marine vegetation with the deep water imaging of the blue band . Your eyes can see reflected blue light. Depending on the clarity of water, visible light like the blue band can reach depths up to 20-30 meters. Tweak your Spectral Signature using the blue band for deep water imaging. DEEP WATER IMAGING: Because of its finer level of Spectral penetration in coastal areas, underwater reefs, water turbidity and sediment, submerged aquatic vegetation, turbidity and bathymetric mapping makes for some unique Remote sensing applications.

6 SMOKE PLUMES, ATMOSPHERIC HAZE AND CLOUDS: The visible range blue is shorter wavelengths like blue ( m), which are most sensitive to atmospheric haze. This makes the blue band ideal for detecting smoke plumes because shorter wavelengths more easily scattered by smaller particles. CLOUDS, SNOW AND ROCK: Separating clouds from snow and rock because the Spectral Signature to haze and thin clouds is highly sensitive to both blue and red wavelengths. (Cloud and Snow Discrimination) 3 Green ( m) To many, green signifies nature plants, trees and forest. It s the same in Remote sensing. The green portion covers reflectance peak from leaf surfaces (hence the color green that we see). This also means the Spectral reflectance is low in the blue and red regions of the spectrum because of the absorption by chlorophyll during photosynthesis. PLANT VIGOR AND VEGETATION: Green is used to discriminate broad vegetation classes and discriminating plant material.

7 A Spectral Signature of reflectance peak from leaf surfaces is highlighted in the green band . ALGAL AND cyanobacterial BLOOMS: Clear water generally reflects little in any part of the spectrum. However, reflectance is maximum at the blue end of the spectrum hence the color blue our eyes see. But when algal blooms are present, green is where the maximum reflectance generally occurs. URBAN RECREATION: The grassy and forested greens in urban areas are indicated of recreational areas like parks, golf courses and cemeteries. 4 Red ( m) Tropical soils, the built environment and geological features generally have red contribute to its Spectral Signature . Red is used in formulas like the Normalized Difference Vegetation Index (NDVI) because how much plants absorb red (and blue) light. SOIL TYPES AND GEOLOGIC FEATURES: The reflectance of bare soil generally depends on its composition. The soils in Australia almost mimics Mars because of its redness.

8 Soils rich in iron-oxide with its rust color has a high reflectance in the red Spectral band great for this type of soil Spectral Signature . (Predicting soil properties) BUILT AND NATURAL ENVIRONMENT: The red band has been used for discriminating between man-made objects and vegetation. A Spectral Signature for cultural features such as roads and human settlements are best detected in the red band . CHLOROPHYLL ABSORPTION: One-half of the variables used in NDVI healthy vegetation reflects more near-infrared (NIR) and green light compared to other wavelengths. It absorbs more red and blue light and is used to quantify vegetation in NDVI. 5 Yellow ( m) The yellow band is relatively new and can be found as early as Worldview-2. This band collects a bit finer detail from the m portion of the spectrum. Simply, it collects the yellowness of a particular object. INSECT DISEASE IN TREE CROWNS: One study developed a yellowness index (YI) as a measure for chlorosis of leaves in stressed plants and manganese-deficient soybean leaves.

9 (Estimating stressed vegetation with the Yellowness Index) FEATURE CLASSIFICATION: The yellow band can be useful in different seasons to delineate invasive grass and other general features. It s also been used to classify individual tree species and crop types by season. 6 Red-Edge ( m) The red-edge band is between the near infrared and red band . For chlorophyll, the red portion strongly absorbs light while the near infrared generates a strong reflection. In the transition between these two Spectral Bands is the red-edge band . PLANT HEALTH AND AGE STATUS: With a high reflectivity of plant health and vitality, the vegetation response from the red-edge is often greatest for both the chlorophyll content and the leaf structure such as leaf area index (LAI). (Identifying plant stress using red-edge) CROP MONITORING: Applying the red-edge band to precision farming and even discriminating between healthy and crops impacted by disease. It s been used to distinguish between crop types and nutrition.

10 7 Near Infrared 1 NIR-1 ( m) Near-infrared reflectance is one of the most powerful ways to classify healthy vegetation. Separating classes like water and vegetation is always easier in the near-infrared region. This is because in the near infrared spectrum, healthy plants reflect it, while near infrared radiation is absorbed more by water. BIOMASS CONTENT: The internal structure of healthy chlorophyll uncannily reflects near-infrared radiation. But when plants wilt and eventually die, the near-infrared response lessens until nothing is left. ARCHAEOLOGICAL SITES: Unearthing ancient archaeological sites by interpreting denser mud bricks, crop marks and subtle differences in vegetation, soil, and geology from near-infrared radiation. (Near-infrared aerial crop mark archaeology) NORMALIZED DIFFERENCE VEGETATION INDEX (NDVI): Near-infrared is used in Spectral signatures for measuring plant health. This can be done using the NDVI ratio, consisting of near-infrared and red Bands .