Spectral resolution

Geography glossary

Topic: Remote Sensing, GIS and GNSS · NCERT: Class 11, Ch 6 "Introduction to Remote Sensing"

Meaning

Spectral resolution is how many wavelength bands (narrow slices of light) a sensor records, and how narrow each band is. More bands, and narrower ones, help a sensor tell apart surfaces that look the same to the human eye, such as water and dry land, or one crop and another.

Explanation

The idea behind it: dispersion of light (NCERT Box 6.2)

  • White light is a mix. It is made of many wavelengths (colours) together.
  • A prism or a rainbow splits it.
  • A glass prism, or raindrops in a rainbow, bend each wavelength by a slightly different amount.

  • So white light spreads out into separate colours (VIBGYOR).

  • A multispectral sensor does the same thing. A multispectral sensor is one that records several bands at once.

  • It splits the energy reflected from the ground into bands.
  • It records each band separately, as its own image.

How bands help: spectral signature

  • Every surface reflects light in its own way.
  • Each surface reflects a different amount of energy in each band.

  • This pattern is called its spectral signature, a kind of fingerprint made of light.

  • More bands show more of that fingerprint.

  • With only a few wide bands, two surfaces can look alike.

  • With many narrow bands, small differences show up, so the surfaces can be told apart.

  • Comparing bands side by side separates surfaces. A surface that is dark in one band and bright in another is easy to pick out.

What sets it and how it is measured

  • Measure: the number of bands and how narrow each band is.
  • Set by: the detector design of the sensor, not the height of the orbit.
  • Finer spectral resolution means more bands and narrower bands.
  • Band regions used in practice:
  • VNIR (visible and near-infrared): light we can see, plus infrared just beyond red.
  • SWIR (short-wave infrared): useful for mapping soil moisture and crop water stress.

The trade-off

  • A narrow band collects less energy.
  • Less light reaches the detector in each band.

  • So the sensor needs bigger pixels or more sensitive detectors to make up for it.

  • This is why panchromatic images are sharper.

  • Panchromatic means one wide black-and-white band.
  • One wide band collects more light, so each pixel can be made smaller.
  • So a satellite's panchromatic image is always sharper than its multispectral images. The multispectral images have finer spectral detail but coarser spatial detail.

Where it occurs

  • Najafgarh, Delhi (NCERT Fig. 6.11, Resourcesat-1, 2005):
  • Fresh water absorbs energy in band 4 (infrared), so water bodies look dark in band 4.
  • Dry surfaces reflect in band 2 (green), so dry land looks bright in band 2.

  • Putting the two bands side by side clearly separates water from dry land.

  • Resourcesat-2A (India, PSLV-C36, 7 December 2016): [2]

  • LISS-4: three bands in the VNIR [2].
  • LISS-3 and AWiFS (Advanced Wide Field Sensor): three VNIR bands plus one SWIR band each [2].

  • The SWIR band helps map soil moisture and crop water stress over Indian farmland.

  • Landsat 8 (USA, launched 11 February 2013): it records 11 spectral bands across two instruments, OLI (Operational Land Imager) and TIRS (Thermal Infrared Sensor) [3].

  • How India uses it: multi-band data from the Resourcesat series is used to map water bodies, crops, soil moisture and crop stress. This data is shared through Bhuvan and opened up under the National Geospatial Policy 2022.

Don't confuse with

  • Spatial resolution: this is the ground size of one pixel (for example, IRS PAN 5.8 m), so it answers "how small an object can it see?" Spectral resolution answers "how many colours (bands) can it tell apart?" Questions often mix up the two.
  • Radiometric resolution: this counts grey levels, the number of brightness steps within a single band (for example, Landsat 8 OLI is 12-bit, which gives 4,096 values [3]). Spectral resolution counts how many bands there are, not the brightness steps inside each band.
  • Temporal resolution: this is the revisit interval and is set by the orbit (Landsat 8 alone: 16 days [3]). Spectral resolution is set by the detector design.
  • Panchromatic image: this is one wide black-and-white band, so it has the finest spatial detail but the poorest spectral detail. A multispectral image has many bands and finer spectral resolution.

Prelims Hooks

  • Spectral resolution is based on the dispersion of light, as seen in a prism and a rainbow (NCERT Box 6.2).
  • Najafgarh, Delhi (Resourcesat-1, 2005): fresh water absorbs in band 4 (infrared), and dry surfaces reflect in band 2 (green).
  • Landsat 8 has 11 spectral bands across OLI and TIRS [3].
  • Resourcesat-2A: LISS-3 and AWiFS carry 3 VNIR + 1 SWIR bands, and LISS-4 carries 3 VNIR bands [2].
  • Measure: number of bands and how narrow they are. Set by: detector design. The orbit does not set it.
  • Trap: "Finer spectral resolution means smaller pixels" is wrong. Narrow bands collect less energy, so they usually need bigger pixels. This is why the panchromatic band is the sharpest.

Mains Points

  • Spectral resolution supports resource management (GS-III):
  • Many narrow bands bring out each surface's spectral signature.
  • This helps map water bodies (Najafgarh), soil moisture and crop water stress (SWIR bands on Resourcesat-2A [2]).

  • It supports drought warning, irrigation planning and crop estimates.

  • One sensor cannot be best at everything, so India uses a mix:

  • Narrow bands collect less light, so spectral detail is traded against spatial detail.
  • So planners combine sharp panchromatic images with multispectral images from series such as Resourcesat and Cartosat.

  • The data reaches users through Bhuvan and the National Geospatial Policy 2022.

  • India's multi-band sensors show its growing self-reliance in space:

  • India moved from the early LISS cameras (IRS-1A, 1988) to Resourcesat-2A, which adds a SWIR band to its VNIR bands [2].
  • This gives India its own data for agriculture, water and disaster management, so it depends less on foreign imagery.

Related concepts

Read more

Sources

  1. 1Class 11, Ch 6 "Introduction to Remote Sensing" (primary)
  2. 2ISRO — RESOURCESAT-2A mission pageisro.gov.in · tier 1
  3. 3NASA Science — Landsat 8 missionscience.nasa.gov · tier 2