Send the link below via email or IMCopy
Present to your audienceStart remote presentation
- Invited audience members will follow you as you navigate and present
- People invited to a presentation do not need a Prezi account
- This link expires 10 minutes after you close the presentation
- A maximum of 30 users can follow your presentation
- Learn more about this feature in our knowledge base article
Atmospheric correction (remote sensing)
Transcript of Atmospheric correction (remote sensing)
Case studies from the literature
Atmospheric contributions to ToA Radiance
of satellite imagery
At satellite radiance (or reflectance) includes contributions from the transmission of radiation (see for example Shaepman-Strub et al.)
Contributions the atmospheric path and from adjacent regions can be scattered into the sensors IFOV (resulting in an anomalous registration/pixel value).
In this example taken from Mather (2004), the at satellite signal includes components from the atmosphere (S1 and S2) and from the adjacency effect (Q) as well as the imaged target (P).
According to Mather (2004) at satellite radiance (Ls) is the product of the total downwelling radiation (Htot), the target reflectance , atmospheric transmittance T, and the atmospheric path radiance (Lp).
Shaepman-Strub et al. (2006; eq. 27 & 28) suggest we break down Ls into direct radiance, diffuse radiance and atmospheric radiance components. They note the diffuse component is itself complex and includes atmospheric scattering components.
See also Jones and Vaughan (2019, Box 6.1, p. 138).
Basic Corrections: empirical line method
A dark and a light target are selected and their at-ground reflectance measured. This is plotted against at-satellite radiance and a linear regression is plotted resulting in:
where s is the slope of the regression, and a is the intercept.
a is considered to be the atmospheric component.
Basic Corrections: Dark Object Subtraction
Dark targets such as deep water are identified: these are assumed to have near-zero radiance. Thus the minimum value associated with a dark target is assumed to be the atmospheric component of at-satellite radiance and is subtracted from the pixel values. This is done for each band.
An alternative method plots pixels in the NIR against each band in turn. A linear regression is performed and the offset between NIR and other bands is assumed to be an experession of atmospheric radiance.
Radiative transfer models
These numerical approaches attempt to quantify atmospheric contributions to at-satellite radiance. RT models assume a homogenous atmosphere of given parameters or may be complex layered models of the atmosphere in which Rayleigh or Mie scattering is assumed.
For an example you can try out online see:
after Mather (2004)
When and how to correct atmospheric effects (Song et al., 2001)
Song et al. (2001) identify a need for robust yet simple correction methods. They note that while accurate RT models require accurate, time-sensitive input data that is rarely available.
They identify a need for atmospheric correction "where a common radiometric scale is assumed among ...multi-temporal images" (Song et al., 2001, p.233).
They tested a range of empirical and semi-empirical atmospheric corrections.
Relative atmospheric correction of Landsat TM data
(Song et al. 2001)
Dark Object Subtraction (DOS) and variations on the Dense Dark Vegetation (DDV) approach were tested.
The DDV assumes that dense vegetation acts as a Dark Object in TM bands 1 and 3 whilst TM7 is insensitive to atmospheric effects. Relationships can then be established between the bands to correct for atmospheric effects.
A 'Ridge Method' using 2-D plots of multi-temporal imagery band by band was also analysed. Marked deviations from the linear intercept are viewed as atmospheric artefacts.
Song et al. (2001) found that the simpler methods generally outperformed the more complex approaches. All methods yielded improved results when compared with uncorrected data.
[see their conclusions for a fuller explanation of the implications]
Correction of MODIS visible-MIR reflectance
Vermote et al. (2002)
MODIS surface reflectance data are corrected for aerosol and water vapour effects.
The correction uses a 6S radiative transfer model with consideration of surface and atmosphere BRDFs. The algorithm therefore requires atmospheric aerosol and water vapour data as input (MODIS produces such data).
Validation against AERONET sites and corrected Landsat ETM+ data suggest high accuracy.
Vermote et al. (2002; p.98-9)
MODIS Visible-MIR ToA Reflectance
Vermote et al. (2002)
Jones, H.G. and Vaughan, R.A. (2010) Remote sensing of vegetation: principles, techniques and applications. Oxford university Press.
Mather, P. (2004). Computer processing of remotely-sensed images: An Introduction (3rd Ed.). Wiley.
Rees, W.G. (2001) Physical principles of remote sensing (2nd Ed.). Cambridge University Press.
Shaepman-Strub, G. et al. (2006) Reflectance quantities in optical remote sensing- definitions and case studies. Remote Sens. Environ., 103.
Song, C. et al. (2001) Classification and change detection using Landsat TM data: When and how to correct atmospheric effects. Remote Sens. Environ., 75.
Vermote, E.F. et al. (2002) Atmospheric correction of MODIS data in the visible to middle infrared: first results. Remote Sens. Environ., 83.
Vermote, E.F. (2013) MODIS Land Surface Reflectance http://modis-sr.ltdri.org/ retrieved 2013-03-05.