• Indexs: Computing of indices of vegetation, snow, water, etc

Accés a aquest text d'ajuda com a pàgina web: Indexs

Presentation and options

This application allows to easily calculate vegetation indices and other indices of interest in remote sensing. In case the index of interest is not available in this program, it can be calculated through the CalcImg application. It is also possible to use the CalcImgto rescale the output values.

Indices are spectral band combinations that enhance the contribution of certaincomponents (typically vegetation) observed in the spectral response of a surface, minimizing other effects such as soil, illumination or atmosphere conditions, that can cause interference in the radiometric signal.

The use of quotients and indices to discriminate vegetation and its characteristics is based on the peculiar radiometric behavior of vegetation. The advantage of indices in front of the use of single spectral bands is that they show a better correlation with ecological and agronomic parameters such as biomass, leaf area index (LAI), etc.

Currently this program has implemented the following indices [more detailed explanations of them can be found in the work of Pons and Arcalís (2012) which can be consulted through the Additional references]:

Implemented vegetation indices

DVI (differenced vegetation index):

Vegetation index obtained from the difference in reflectance between the near infrared and red spectral regions of a sensor.


DVI= 2.4 · ρ_NIR - ρ_R
where:
ρ_NIR is the reflectance in the near infrared spectral region
ρ_R is the reflectance in the red spectral region.
Richardson, A.J., Wiegand, C.L. (1977) Distinguishing Vegetation from Soil Background Information. Photogrammetric Engineering and Remote Sensing, 43(12):1541-1552. https://www.asprs.org/wp-content/uploads/pers/1977journal/dec/1977_dec_1541-1552.pdf

RVI (ratio vegetation index):

Vegetation index obtained from the simple quotient of near infrared and red reflectance that has been used for biomass and leaf area index (LAI) assessments since 1969. It was originally named vegetation index number (VIN) (Pearson and Miller 1972).


RVI= ρ_NIR / ρ_R
where:
ρ_NIR is the reflectance in the near infrared band
ρ_RB is the reflectance in the red band.

The RVI takes values between 0 and infinity, with higher values indicating more vegetation coverage. It was, along with the GVI (see Pons and Arcalís, 2012, for details), one of the first vegetation indices proposed.

Pearson, R.L., Miller, L.D. (1972) Remote mapping of standing crop biomass for estimation of the productivity of the shortgrass prairie, Pawnee National Grasslands, Colorado. Proceedings of the 8th International Symposium on Remote Sensing of the Environment II:1355-1379.

NDVI (normalized difference vegetation index):

Vegetation index used to measure vegetation vigor based on reflectance in the red and near infrared spectral regions. This vegetation index has been the most used in remote sensing of vegetation in the last decades. It was first used by Kriegler et al. (1969) and developed by Rouse et al. (1973) from the analysis of Landsat-MSS data. They found that, although the simple quotient (RVI) could be used as a measure of relative greenness, the position and cyclical deviations could cause important errors. They defined the normalized difference vegetation index (NDVI) in the following way:


NDVI= (ρ_NIR - ρ_R) / (ρ_NIR + ρ_R)where:
ρ_NIR is the reflectance in the near infrared spectral region
ρ_R is the reflectance in the red spectral region.

An interesting aspect of this index regarding the simple quotient is that it has values between -1 and 1 as long as the variables are expressed in the same "units" (in fact, in the same range of variation), making its interpretation easier.

This index has been used to estimate a large number of variables: leaf area index (LAI), net CO₂ flux, photosynthetically active radiation absorbed by the plant (APAR), vegetation productivity, chlorophyll content, phenological dynamics, potential evapotranspiration, amount of rain-fall on the vegetation coverage, etc. The most common applications are: global monitoring of vegetation coverage, deforestation studies, desertification, biomes characterization on a global scale and drought studies and forest fire risk from multitemporal series of images.

Kriegler, F.J., Malila, W.A., Nalepka, R.F., Richardson, W. (1969) Preprocessing transformations and their effects on multispectral recognition. Proceedings of the Sixth International Symposium on Remote Sensing of Environment, 97-131.

Rouse, J.W., Haas, R.H., Schell, J.A., Deering, D.W. (1973) Monitoring vegetation systems in the great plains with ERTS. Third ERT Symposium, NASA SP-351, I:309-317.https://ntrs.nasa.gov/api/citations/19740022614/downloads/19740022614.pdf

Rouse, J.W., Haas, R.H., Deering, D.W., Schell, J.A. (1973) Monitoring the vernal advancement and retrogradation (green wave effect) of natural vegetation. NASA/GSFCT Type II Report. Remote Sensing Center, Texas A&M University, Greenbelt, Maryland. Page 76. https://core.ac.uk/download/pdf/80640125.pdf

TVI (transformed vegetation index):

Vegetation index created from a transformation of the NDVI index that avoids working with negative values.


TVI= √(NDVI + 0.5)
where:
NDVI is the normalized difference vegetation index. In the MiraMon implementation, NDVI values in the range [-1,0.5] are mapped to 0.
Rouse, J.W., Haas, R.H., Schell, J.A., Deering, D.W. (1973) Monitoring vegetation systems in the great plains with ERTS. Third ERT Symposium, NASA SP-351, I:309-317.https://ntrs.nasa.gov/api/citations/19740022614/downloads/19740022614.pdf

Rouse, J.W., Haas, R.H., Deering, D.W., Schell, J.A. (1973) Monitoring the vernal advancement and retrogradation (green wave effect) of natural vegetation. NASA/GSFCT Type II Report. Remote Sensing Center, Texas A&M University, Greenbelt, Maryland. Page 76. https://core.ac.uk/download/pdf/80640125.pdf

PVI (perpendicular vegetation index):

Vegetation index based on the fact that the NIR and red reflectance vary together, and in the opposite direction, with increasing vegetation density, and that these variations mean that a point located above the soil baseline in which the amount of vegetation increases is separated from the line in an upward and towards the left direction.


PVI= (ρ_NIR - γ·ρ_R - δ) / √(γ² + 1)
where:
ρ_NIR is the reflectance in the near infrared spectral region
ρ_R is the reflectance in the red spectral region
γ and δ are the slope and the ordinate at the origin, respectively, of the soil baseline.

The result can take very different values, and therefore it is difficult to establish specific margins, although as an approximation it can be said that negative PVI values correspond to water areas, PVI values around zero correspond to bare soil and positive PVI values correspond to areas of active vegetation.

Some limitations of this index are the fact that it can vary depending on the observation geometry and that it is sensitive to atmospheric scattering.

Richardson, A.J., Wiegand, C.L. (1977) Distinguishing Vegetation from Soil Background Information. Photogrammetric Engineering and Remote Sensing, 43(12):1541-1552. https://www.asprs.org/wp-content/uploads/pers/1977journal/dec/1977_dec_1541-1552.pdf

WDVI (weighted difference vegetation index):

Vegetation index that modulates the NDII index response (see Pons and Arcalís, 2012, for details).


WDVI= ρ_NIR - γ·ρ_R
where:
ρ_NIR is the reflectance in the near infrared spectral region
ρ_R is the reflectance in the red spectral region
γ is the slope of the soil baseline.

The result can take very diverse values, and therefore it is difficult to establish specific margins, although as an approximation it can be said that negative WDVI values correspond to water areas, values around zero correspond to bare soil and positive values correspond to active vegetation, with the greater the vigor of the vegetation, the greater the value obtained.

Clevers, J.G.P.W. (1989) Application of a weighted infrared-red vegetation index for estimating leaf Area Index by Correcting for Soil Moisture. Remote Sensing of Environment, 29(1):25-37. https://doi.org/10.1016/0034-4257(89)90076-X

SAVI (soil adjusted vegetation index):

Index that intends to minimize the soil effect in the result of the calculation of the vegetation indices explained above, especially on partially vegetated surfaces (dry and arid areas).


SAVI= (1 + L) · (ρ_NIR - ρ_R) / (ρ_NIR + ρ_R + L)
where:
ρ_NIR is the reflectance in the near infrared spectral region
ρ_R is the reflectance in the red spectral region
L is a constant of adjustment of the soil influence that depends on the vegetation coverage and can take a value between 0 and 1. L = 0 is often taken for areas with very high vegetation density, L = 0.5 for areas with intermediate vegetation density and L = 1 for areas with very low vegetation coverage. When L = 0, the SAVI is exactly the same as the NDVI.

The SAVI index presents values between -1 and 1. Negative SAVI values correspond to areas with very low vegetation coverage, with values close to -1 for bare soil or water, while positive values correspond to areas with active vegetation, with more vigor of the vegetation the closer to 1 is the value obtained.

Huete, A.R. (1988) A soil-adjusted vegetation index (SAVI). Remote Sensing of Environment, 25(3):295-309. https://doi.org/10.1016/0034-4257(88)90106-X

TSAVI (transformed soil adjusted vegetation index):

Variation of the SAVI index to consider the effect of the soil brightness.


TSAVI= γ · (ρ_NIR - γ · ρ_R - δ) / [ρ_R + γ · ρ_NIR - γ · δ + κ · (1 + γ²)]
where:
ρ_NIR is the reflectance in the near infrared spectral region
ρ_R is the reflectance in the red spectral region
γ and δ are, respectively, the slope and the ordinate at the origin of the soil baseline
κ represents a fitting factor to minimize the soil effect, which the authors recommend applying with a value of 0.08.

The TSAVI takes values between 0 and 1. TSAVI values close to zero correspond to bare soil and values close to 1 (generally greater than 0.70) correspond to areas with active vegetation.

This index is very suitable for semi-arid regions, where soil moisture and vegetation coverage are very low.

Baret, E., Guyot, G., Major, D. J. (1989) A vegetation index which minimizes soil brightness effects on LAI and APAR estimation. Proceedings of the 12th Canadian Symposium on Remote Sensing, Vancouver, Canada, 1355-1358. https://doi.org/10.1109/IGARSS.1989.576128

Baret, E., Guyot, G. (1991) Potentials and limits of vegetation indices for LAI and APAR assessment. Remote Sensing of Environment, 35:161-173. https://doi.org/10.1016/0034-4257(91)90009-U

MSAVI-1 (modified soil adjusted vegetation index - 1):

Variation of the SAVI index, with a fitting factor L that is not constant, but is a function that varies inversely with the presence of vegetation. The adjustment factor "L" of SAVI index is not constant, but an inverse function of the vegetation coverage. There have been proposed two solutions:


MSAVI= (1 + L) · (ρ_NIR - ρ_R) / (ρ_NIR + ρ_R + L)
where:
ρ_NIR is the reflectance in the near infrared spectral region
ρ_R is the reflectance in the red spectral region
L is an adjustment parameter of the soil influence, a factor that depends on γ which is the slope of the soil baseline, on the NDVI and on the WDVI:
L = 1 - 2 · γ · NDVI · WDVI
L can take values between 0 and 1. Values close to 0 correspond to areas with high vegetation density, while values close to 1 correspond to areas with very low vegetation coverage. When L = 0, MSAVI-1 is exactly equal to NDVI.

MSAVI-1 takes values between -1 and 1. MSAVI-1 negative values correspond to areas with very low vegetation coverage, values close to -1 correspond to bare soils, while positive values correspond to areas with active vegetation, with more vigor of the vegetation the closer to 1 the value obtained.

Qi, J., Chehbouni, A., Huete, A.R., Kerr, Y.H., Sorooshian, S. (1994). A modified soil adjusted vegetation index. Remote Sensing of Environment, 48:119-126. https://doi.org/10.1016/0034-4257(94)90134-1

MSAVI-2 (modified soil adjusted vegetation index - 2):

Variation of the SAVI index.


MSAVI-2= (2 · ρ_NIR + 1 - √[(2 · ρ_NIR + 1)² - 8 (ρ_NIR - ρ_R)])/2
where:
ρ_NIR is the reflectance in the near infrared spectral region
ρ_R is the reflectance in the red spectral region.

The result can take very different values and, therefore, it is difficult to establish specific values, although as an approximation it can be said that the MSAVI-2 takes negative values in areas with vegetation coverage very low and positive in areas with active vegetation, with more vigor of the vegetation the closer to 1 is the value obtained.

Qi, J., Chehbouni, A., Huete, A.R., Kerr, Y.H., Sorooshian, S. (1994). A modified soil adjusted vegetation index. Remote Sensing of Environment, 48:119-126. https://doi.org/10.1016/0034-4257(94)90134-1

TWVI (two-axis adjusted vegetation index):

Vegetation index that improves some aspects of the PVI and the SAVI index by studying the complexity of the soil baseline, specially when the distance D between a pixel of soil without vegetation and the soil baseline is large.


TWVI= (1 + L)·[(ρ_NIR - ρ_R - Δ) / (ρ_NIR + ρ_R + L)]
where:
Δ= √ 2 · e^(-K·LAI)· D
and:
D= (ρ_{NIR_soil} - γ · ρ_{R_soil} - δ) / √ (1 + γ²)
where:
ρ_NIR is the reflectance in the near infrared spectral region
ρ_R is the reflectance in the red spectral region
ρ_{NIR_soil} is the reflectance of the soil in the near infrared spectral region
ρ_{R_soil} is the reflectance of the soil in the red spectral region
γ and δ are the slope and the ordinate at the origin, respectively, of the soil baseline,
L is an adjustment constant of the soil influence that can take a value between 0 and 1. A value close to 0 corresponds to areas with high vegetation density, while a value close to 1 corresponds to areas with very low vegetation coverage,
K is an extinction coefficient related to the spread angle of the leaves,
LAI is the leaf area index.

The TWVI requires preliminary knowledge of the study area, in particular of LAI and soil reflectances.

Xia, L. (1994) A two-axis adjusted vegetation index (TWVI). International Journal of Remote Sensing, 15(7):1447-1458. https://doi.org/10.1080/01431169408954176

ARVI (atmospherically resistant vegetation index):

Vegetation index obtained from the improvement of the NDVI against the effect of atmospheric absorption and scattering. To achieve this, the red band is substituted by a red-blue band that considers the different atmospheric effects in these two spectral regions. These effects are more important at pixels far from nadir and in the case of oblique observations, as is the case at the lateral ends of images from MODIS sensors and other in sensors with a large FOV.


ARVI= (ρ_NIR - ρ_RB) / (ρ_NIR + ρ_RB)
where:
ρ_NIR is the apparent reflectance (at the top of the atmosphere) in the near infrared spectral region
ρ_RB is the apparent reflectance in the combined band of the red and blue bands, according to the formula:
ρ_RB= ρ_R - η·(ρ_B - ρ_R)

where:
ρ_R is the apparent reflectance in the red spectral region
ρ_R is the apparent reflectance in the blue spectral region
η (eta, written γ in the original article) is an atmospheric corrector that when is 0 causes that the ARVI is equal to the NDVI; original research by Kaufman and Tanré (1992) showed that η = 1 is a suitable compromise for most terrestrial remote sensing applications, especially those focused on vegetation in atmospheres with small and moderate aerosol particle sizes (for example, anthropogenic aerosols and smoke) and in environments with large particles, but very dry conditions (Sahel), and that values close to 2 are not suitable.

This index takes values between -1 and 1. Negative values of the index correspond to areas of water, values close to zero correspond to bare soil and positive values correspond to areas of vegetation, with more vigor of the vegetation the closer to 1 is the value obtained.

Optionally, it is possible to indicate the treatment of the pixels of the RB band calculated during the process when these are outside the range [0,1]. The treatment can be "Assign NoData" (default value, which can be specified with /RB_CRITERI=ND on the command line), "Bound between [0,1]" (indicated with /RB_CRITERI =LIM on the command line), "Allow free calculation" (indicated with /RB_CRITERI=IGNOR on the command line).

It is also possible to specify the treatment of the pixels of the obtained index when they are outside the range [-1,1]. The treatment can be "Assign NoData" (default value, which can be specified with /IDX_CRITERI=ND on the command line), "Bound between [0,1]" (indicated with /IDX_CRITERI =LIM on the command line), "Allow free calculation" (indicated with /IDX_CRITERI=IGNOR on the command line).

Important: this index, since it works with apparent reflectances (or even in apparent radiances), should not be applied to images that have been radiometrically corrected to obtain object-level reflectances (at the bottom of the atmosphere).

Kaufman, Y.J., Tanré, D. (1992) Atmospherically resistant vegetation index (ARVI) for EOS-MODIS. IEEE Trans. on Geosciences and Remote Sensing, 30(2):261-270. https://doi.org/10.1109/36.134076

SARVI (soil adjusted and atmospherically resistant vegetation index):

Vegetation index that corrects the radiance of the red band by the effect of aerosols incorporating the blue band, and introduces another correction to minimize the soil contribution. It is a combination of the ARVI and SAVI indices.


SARVI= (1 + L) · (ρ_NIR - ρ_RB) / (ρ_NIR + ρ_RB + L)
where:
ρ_NIR is the apparent reflectance (at the top of the atmosphere) in the near infrared spectral region
ρ_RB is the combined apparent reflectance of the two spectral regions of red and blue, which is calculated:ρ_RB= ρ_R - η·(ρ_B- ρ_R)

where:
ρ_R is the apparent reflectance in the red spectral region
ρ_B is the apparent reflectance in the blue spectral region
η (eta, written γ in the original article) is an atmospheric corrector that when is 0 causes that the ARVI is equal to the NDVI; original research by Kaufman and Tanré (1992) showed that η = 1 is a suitable compromise for most terrestrial remote sensing applications, especially those focused on vegetation in atmospheres with small and moderate aerosol particle sizes (for example, anthropogenic aerosols and smoke) and in environments with large particles, but very dry conditions (Sahel), and that values close to 2 are not suitable

L is an adjustment constant of the soil influence that can take a value between 0 and 1. Often L = 0 is taken for areas with very high vegetation density, L = 0.5 for areas with intermediate density and L = 1 for areas with very low vegetation density. When L = 0, SARVI is exactly equal to ARVI.

This index takes values between -1 and 1. Negative values of the index correspond to areas of water, values close to zero correspond to bare soil and positive values correspond to areas with vegetation, with more vigor of the vegetation the closer to 1 the value obtained.

Optionally, it is possible to indicate the treatment of the pixels of the RB band calculated during the process when these are outside the range [0,1]. The treatment can be "Assign NoData" (default value, which can be specified with /RB_CRITERI=ND on the command line), "Bound between [0,1]" (indicated with /RB_CRITERI =LIM on the command line), "Allow free calculation" (indicated with /RB_CRITERI=IGNOR on the command line).

It is also possible to specify the treatment of the pixels of the obtained index when they are outside the range [-1,1]. The treatment can be "Assign NoData" (default value, which can be specified with /IDX_CRITERI=ND on the command line), "Bound between [0,1]" (indicated with /IDX_CRITERI =LIM on the command line), "Allow free calculation" (indicated with /IDX_CRITERI=IGNOR on the command line).

Important: this index, since it works with apparent reflectances (or even in apparent radiances), do not must be applied to images that have been radiometrically corrected to obtain object-level reflectances (at the bottom of the atmosphere).

Kaufman, Y.J., Tanré, D. (1992) Atmospherically resistant vegetation index (ARVI) for EOS-MODIS. IEEE Trans. on Geosciences and Remote Sensing, 30(2):261-270. https://doi.org/10.1109/36.134076

TSARVI (transformed soil atmospherically resistant vegetation index):

This index is a variant of the SARVI index that considers the effect of the brightness of the soil.


TSARVI= γ_RB·(ρ_NIR - γ_RB · ρ_RB - δ_RB) / [γ_RB· ρ_NIR + ρ_RB - γ_RB · δ_RB + κ·(1 + γ_RB²)]
where:
ρ_NIR is the apparent reflectance (at the top of the atmosphere) in the near infrared spectral region
ρ_RB is the combined apparent reflectance of the two spectral regions of red and blue, which is calculated:
ρ_RB= ρ_R - η·(ρ_B- ρ_R)

where:
ρ_R is the apparent reflectance in the red spectral region
ρ_B is the apparent reflectance in the blue spectral region
η (eta, written γ in the original article) is an atmospheric corrector that when is 0 causes that the ARVI is equal to the NDVI; original research by Kaufman and Tanré (1992) showed that η = 1 is a suitable compromise for most terrestrial remote sensing applications, especially those focused on vegetation in atmospheres with small and moderate aerosol particle sizes (for example, anthropogenic aerosols and smoke) and in environments with large particles, but very dry conditions (Sahel), and that values close to 2 are not suitable.

γ_RB and δ_RB are, respectively, the slope and ordinate at the origin of the soil baseline in the red-blue spectral space and NIR, defined by
ρ_NIR = γ_RB · ρ_RB + δ_RB
Note: the parameter γ of TSARVI is not the parameter η used to compute the RB band of the ARVI, SARVI and TSARVI indices.
κ represents a fitting factor to minimize ground noise, which the authors recommend applying a value of κ = 0.08.

This index takes values between -1 and 1. Negative values of the index correspond to areas of water, values close to zero correspond to bare soil and positive values correspond to areas of vegetation, with more vigor of the vegetation the closer to 1 is the value obtained.

Optionally, it is possible to indicate the treatment of the pixels of the RB band calculated during the process when these are outside the range [0,1]. The treatment can be "Assign NoData" (default value, which can be specified with /RB_CRITERI=ND on the command line), "Bound between [0,1]" (indicated with /RB_CRITERI =LIM on the command line), "Allow free calculation" (indicated with /RB_CRITERI=IGNOR on the command line).

It is also possible to specify the treatment of the pixels of the obtained index when they are outside the range [-1,1]. The treatment can be "Assign NoData" (default value, which can be specified with /IDX_CRITERI=ND on the command line), "Bound between [0,1]" (indicated with /IDX_CRITERI =LIM on the command line), "Allow free calculation" (indicated with /IDX_CRITERI=IGNOR on the command line).

Important: this index, since it works with apparent reflectances (or even in apparent radiances), should not be applied to images that have been radiometrically corrected to obtain object-level reflectances (at the bottom of the atmosphere).

Bannari, A., Morin, D., He, D.C. (1994) High spatial and spectral resolution remote sensing for the management of the urban environment. First International Airborne Remote Sensing Conference and Exhibition, Strasbourg, France, III:247-260.

AVI (angular vegetation index):

Vegetation index based on the angle defined between the reflectances at 555 nm, 670 nm and 870 nm corresponding to the green, red and NIR spectral regions, respectively, of the IRR sensor of the ATSR-2 instrument 2 of the ERS-2, which characterizes the absorption of chlorophyll, to normalize the effects of the soil and the atmosphere and maintain the sensitivity to the properties of the vegetation.


AVI= 2 · (π - (a₁ + a₂)) / π
where:
a₁= tan^-1 [((λ_NIR - λ_R) / λ_R) / (ρ_NIR - ρ_R)]
a₂= tan^-1 [((λ_R - λ_G) / λ_R) / (ρ_G - ρ_R)]
where:
λ_G, λ_R and λ_NIR correspond to the centers of the green spectral regions, the red and the near infrared, respectively, which are obtained from the metadata
ρ_G, ρ_R and ρ_NIR correspond to the reflectances in the spectral regions of the green, red and near infrared, respectively.
The tan^-1 function returns the angle in radians. According to the authors, the data are scaled between 0 and 1.
Plummer, S.E., North, P.R., Briggs, S.A. (1994) The angular vegetation index: an atmospherically resistant index for the second along track scanning radiometer (ATSR-2). Proceedings of the Sixth International Symposium Physical Measurements and Signatures in Remote Sensing, 717-722, Val d'Isère, France.

GEMI (global environment monitoring index):

This index was proposed for NOAA-AVHRR sensor to have a global knowledge of the vegetation on the Earth. It is a non-linear index that minimizes the soil and atmospheric effects.


GEMI= η · (1 - 0.25·η) - (ρ_R - 0.125) / (1 - ρ_R)
where:
η= [2· (ρ²_NIR - ρ²_R) + 1.5 · ρ_NIR + 0.5· ρ_R] / (ρ_NIR + ρ_R + 0.5)
where
ρ_NIR is the reflectance in the near infrared spectral region
ρ_R is the reflectance in the red spectral region.
Pinty, B., Verstraete, M.M. (1992) GEMI: a non-linear index to monitor global vegetation from satellites. Vegetatio, 101:15-20. https://doi.org/10.1007/BF00031911

NDWI-MF (normalized difference water index - McFeeters):

The NDWI was created by McFeeters (1996) to delineate open-water features such as lakes and reservoirs, and to improve their detection in satellite images.


NDWI-MF= (ρ_G - ρ_NIR) / (ρ_G + ρ_NIR)
where:
ρ_G is the reflectance in the green spectral region.
ρ_NIR is the reflectance in the near infrared spectral region.

In this index water surfaces have positive values, while land and vegetation surfaces have 0 or negative values.

This index corresponds to NDWI¹ of Diccionari terminològic de teledetecció by Pons and Arcalís (2012).

McFeeters, S.K. (1996) The use of the normalized difference water index (NDWI) in the delineation of open water features. International Journal of Remote Sensing, 17:1425-1432. https://doi.org/10.1080/01431169608948714

NDWI-OT (normalized difference water index - Ouma & Tateishi):

The NDWI-OT index is another of the very diverse approaches to indices for the delimitation of water bodies, in this case one used by Ouma & Tateishi (2006) as a part of an approach called WI. Specifically, the NDWI-OT (which in the authors' terminology is called NDWI₃), uses the SWIR and NIR bands (bands 5 and 4 of the TM/ETM+ sensors) to quickly delineate the coastline of water bodies.


NDWI-OT= (ρ_NIR - ρ_SWIR) / (ρ_NIR + ρ_SWIR)
where:
ρ_NIR is the reflectance in the near infrared spectral region
ρ_SWIR is the reflectance in the shortwave infrared region.
Ouma, Y.O., Tateishi, R. (2006) A water index for rapid mapping of shoreline changes of five East African Rift Valley lakes: an empirical analysis using Landsat TM and ETM+ data. International Journal of Remote Sensing, 27:3153-3181. https://doi.org/10.1080/01431160500309934

EVI (enhanced vegetation index):

The EVI index obtains a response from the structural variations of the vegetation canopy including the leaf area index (LAI), the type and architecture of the canopy, etc. The EVI index was developed to optimize the vegetation signal with improved sensitivity for high biomass densities. This optimization is achieved by separating the signal coming from the vegetation to the signal coming from the atmospheric influence.


EVI= G · (ρ_NIR - ρ_R) / (ρ_NIR + C₁ · ρ_R - C₂ · ρ_B + L)
where:
ρ_NIR is the reflectance in the near infrared spectral region
ρ_R is the reflectance in the red spectral region
ρ_B is the reflectance in the blue spectral region
C₁ and C₂ are two correction coefficients for aerosol atmospheric scattering, generally used with values of C₁=6 and C ₂=7.5
L is an adjustment constant of the soil influence, generally used with value L = 1
G is a gain factor, initially set with a value of G=2.0 (Huete et al., 1999), but subsequently generally used with value of G=2.5 (Huete et al., 2002; Glenn et al., 2008). To obtain the index calculated with a value of G=2.0, multiply the resulting raster by 2/2.5, or by 0.8.

This index takes values between 0 and 1, the closer to 1 the greater the vigor of the vegetation.

Since 2000, and following the launch of the two MODIS sensors on the Terra and Aqua satellites, the EVI has been adopted as a standard NASA product and has become very popular among its users for its ability to eliminate background noise, as well as saturation, a characteristic problem of NDVI in areas of high vegetation density.

Glenn, E., Huete, A.R., Nagler, P., Nelson, S. (2008) Relationship Between Remotely-sensed Vegetation Indices, Canopy Attributes and Plant Physiological Processes: What Vegetation Indices Can and Cannot Tell Us About the Landscape. Sensors, 8(4):2136-2160. https://doi.org/10.3390/s8042136

Huete, A.R, Justice, C., van Leeuwen, W. (1999) MODIS Vegetation Index (MOD 13) Algorithm Theoretical Basis Document, Version 3. https://modis.gsfc.nasa.gov/data/atbd/atbd_mod13.pdf

Huete, A.R, Didan, K., Miura, T., Rodriquez, E., Gao, X., Ferreira, L. (2002) Overview of the radiometric and biophysical performance of the MODIS vegetation indices. Remote Sensing of Environment, 83:195?213. https://doi.org/10.1016/S0034-4257(02)00096-2

AFRI2.1 (aerosol free vegetation index 2.1):

The AFRI_2.1 index is developed from NDVI, using the shortwave infrared band (SWIR), 2.1 µm, instead of the red band to reduce the effects of aerosols. It is proposed with the ρ_2.1 MODIS band and the near infrared band, ρ_NIR.


AFRI_2.1= (ρ_NIR - 0.5 · ρ_2.1) / (ρ_NIR + 0.5 · ρ_2.1)
where:
ρ_NIR is the reflectance in the near infrared spectral region
ρ_2.1 is the reflectance in the shortwave infrared spectral region, 2.1 µm, of MODIS.

The superiority of the AFRI is due to the capacity of the SWIR bands to penetrate the atmospheric column, even with the presence of smoke and sulphates. These indices are therefore useful for evaluating vegetation in the presence of smoke, anthropic pollution or volcanic eruptions. In clear sky conditions with low atmospheric influence these indices correlate highly with the NDVI.

Karnieli, A., Kaufman, Y.J., Remer, L., Wald, A. (2001) AFRI Aerosol free vegetation index. Remote Sensing of Environment, 77(1):10-21. https://doi.org/10.1016/S0034-4257(01)00190-0

AFRI1.6 (aerosol free vegetation index 1.6):

The AFRI_1.6 index is developed from NDVI, using the SWIR bands, 1.6 µm, instead of red band to reduce the effects of aerosols. It is proposed with the ρ_1.6 MODIS band and the near infrared band, ρ_NIR.


AFRI_1.6= (ρ_NIR - 0.66 · ρ_1.6) / (ρ_NIR + 0.66 · ρ_1.6)
where:
ρ_NIR is the reflectance in the near infrared spectral region
ρ_1.6 is the reflectance in the shortwave infrared spectral region, 1.6 µm, of MODIS.

The superiority of the AFRI is due to the capacity of the SWIR bands to penetrate the atmospheric column, even with the presence of smoke and sulphates. These indices are therefore useful for evaluating vegetation in the presence of smoke, anthropic pollution or volcanic eruptions. In clear sky conditions with low atmospheric influence these indices correlate highly with the NDVI.

Karnieli, A., Kaufman, Y.J., Remer, L., Wald, A. (2001) AFRI Aerosol free vegetation index. Remote Sensing of Environment, 77(1):10-21. https://doi.org/10.1016/S0034-4257(01)00190-0

PRI (photochemical reflectance index):

The PRI is an index of photosynthetic efficiency of the vegetation useful in the estimation of the carbon balance in the Mediterranean ecosystems.


PRI= (ρ₅₃₁ - ρ₅₇₀) / (ρ₅₃₁ + ρ₅₇₀)
where:
ρ₅₃₁ and ρ₅₇₀ correspond to the reflectance in the wavelengths of 531 nm and 570 nm, respectively. These wavelengths can be slightly modified depending on the sensor, the species, etc.

The Mediterranean evergreen forest presents a very low seasonality in vegetation indices such as NDVI or EVI, despite having a high seasonality of carbon fixation. In these cases, the PRI makes it possible to detect which part of the carbon fixation capacity has become effective.

PRI index is related to the use of radiation, with values between -1 and 1, increasing as photosynthetic efficiency increases.

Gamon, J., Peñuelas, J., Field, C. (1992) A narrow-waveband spectral index that tracks diurnal changes in photosynthetic efficiency. Remote Sensing of Environment, 41:35-44. https://doi.org/10.1016/0034-4257(92)90059-S

NDSI (normalized difference snow index):

Index that allows the detection of the snow coverage.


NDSI= (ρ_G - ρ_SWIR) / (ρ_G + ρ_SWIR)
where:
ρ_G and ρ_SWIR correspond to the reflectance in the spectral region of the green and the shortwave infrared, respectively, since the snow is very reflective in the visible band and absorbing in the shortwave infrared. In the case of TM sensor data, band 5 is used as shortwave band.

NDSI values greater than 0.4 usually correspond to areas with snow coverage, although it is often necessary to adjust the value depending on the image to reduce omission and commission errors.

Several authors recommend applying masks to shady areas and over lakes to avoid confusion with snowy areas.

Dozier, J. (1989) Spectral signature of alpine snow cover from the Landsat Thematic Mapper. Remote Sensing of Environment, 28:9-22. https://doi.org/10.1016/0034-4257(89)90101-6

BI (brightness index):

Index used in the study of soil conditions (color, moisture and structure), especially in arid and semi-arid regions.


BI= √(ρ²_R + ρ²_NIR)
where:
ρ_R is the reflectance in the red spectral region
ρ_NIR is the reflectance in the near infrared spectral region.

As this is not a standardized index, the result is not bounded above by any constants. High BI values indicate higher reflectance, a situation that corresponds to dry, bare, light-colored soils.

The use of this index is included on page 81 of the doctoral dissertation of:

Rapinel, S. (2012) Contribution de la télédétection à l'évaluation des fonctions des zones humides: de l'observation à la modélisation prospective. https://tel.archives-ouvertes.fr/tel-00731169/document.

There is also an index with the same name, but using a different formulation (p. 27 of the printed text), adding to the sum the square of the reflectance in the green band:

Escadafal, R., Bacha, S. (1995) Strategy for the Dynamic Study of Desertification. Proceedings of the ISSS International Symposium, Ouagadougou, Burkina Faso, 6-10:19-34. https://horizon.documentation.ird.fr/exl-doc/pleins_textes/pleins_textes_6/colloques2/010008392.pdf

Finally, there is a similar formulation (equation 2), but with the green and red bands, probably less informative due to the usual high correlation between these bands, in the work of:

Marques, M.J., Alvarez, A., Carral, P., Sastre, B., Bienes, R. (2020) The use of remote sensing to detect the consequences of erosion in gypsiferous soils Water Conservation Research, 8:383-392. https://doi.org/10.1016/j.iswcr.2020.10.001.

Note: The name of this index is often associated to erroneous references or to references that use other approximations (linear combinations such as the brightness of the Tasseled Cap transformation, etc). To avoid confusion and waste of time for MiraMon users, here is a reference that do NOT have to do with the index that is implemented as 24 in this application:

Escadafal, R., Belghith, A., Ben Moussa, H. (1994) Indices spectraux pour la dégradation des milieux naturels en Tunisie aride. 6ème Symposium International Mesures Physiques et Signatures en Télédétection. ISPRS-CNES https://horizon.documentation.ird.fr/exl-doc/pleins_textes/pleins_textes_5/b_fdi_20-21/26946.pdf [this page contains a nonsensical formula and also gives a reference that is also unrelated].

NDWI-Chen (normalized difference water index [vegetation moisture] - Chen, Huang, Jackson):

Index proposed by Chen and collaborators in 2005 to determine the water content in vegetation with data from the MODIS sensor. It is based on the absorption of water in the SWIR and is calculated:


NDWI-Chen= (ρ_NIR - ρ_SWIR) / (ρ_NIR + ρ_SWIR)
where:
ρ_NIR and ρ_SWIR correspond to the reflectance in the spectral region centered at 858 nm and 1640 (or 2130) nm, respectively; formulated this way, the NDWI takes values between slightly negative and 1 depending on the water content of the vegetation.

This index corresponds to the one explained in NDWI² of Diccionari terminològic de teledetecció by Pons and Arcalís (2012). The USGS calls it NDMI (for its acronym in English, this reference can be consulte https://www.usgs.gov/landsat-missions/normalized-difference-moisture-index).

Chen, D., Huang, J., Jackson, T.J. (2005) Vegetation water content estimation for corn and soybeans using spectral indices derived from MODIS near- and short-wave infrared bands. Remote Sensing of Environment, 98:225-236. https://doi.org/10.1016/j.rse.2005.07.008

The Indexs program asks for the names of the files corresponding to the red, infrared and, in some cases, blue bands or other spectral bands. It also may require a set of parameters that depend on the index equation.

The NoData values of the original files are stored as NoData values in the output file.All numerical singularities (negative roots, divisions by zero, etc) are resolved by writing a NoData to the resulting raster, except for the TVI, where it makes sense to assign a zero as it corresponds to a saturated value. For the TVI, if a resulting NoData in case of a negative root is desired, please run the formula in the CalcImg.

This application supports all MiraMon raster formats, as input, whether compressed or not, and generates a "real-RLE" type raster when all the input files are compressed in RLE format, and of "real" (uncompressed) type in all other cases.

Additional references:

Bannari, A., Morin, D., Bonn, F., Huete, A. (1995) A review of vegetation indices. Remote Sensing Reviews, 13:95-120. https://doi.org/10.1080/02757259509532298

Bariou, R., Lecamus, D., Le Henaff, F. (1985) Indices de végétation. Dossiers de télédétection, 2, Presses Universitaires de Rennes.

Haboudane, D., Miller, J.R., Pattey, E., Zarco-Tejadada, P.J., Strachan, I.B. (2004) Hyperspectral vegetation indices and novel algorithms for predicting green LAI of crop canopies: Modeling and validation in the context of precision agriculture. Remote Sensing of Environment, 90:337-352. https://doi.org/10.1016/j.rse.2003.12.013

Jackson, R.D., Huete, A.R. (1991) Interpreting vegetation indices. Preventive Veterinary Medicine, 11:185-200.https://doi.org/10.1016/S0167-5877(05)80004-2

Marco-Dos Santos, G., Meléndez-Pastor, I., Navarro-Pedreño, J., Gómez-Lucas, I. (2019) A Review of Landsat TM/ETM based Vegetation Indices as Applied to Wetland Ecosystems. Journal of Geographical Research, 2(1):35-49. https://doi.org/10.30564/jgr.v2i1.499

Perry, C.R., Lautenschlager, L.F. (1984) Functional equivalence of spectral vegetation indices. Remote Sensing of Environment, 14:169-182. https://doi.org/10.1016/0034-4257(84)90013-0

Pons X, Arcalís A (2012) Diccionari terminològic de teledetecció. Enciclopèdia Catalana i Institut Cartogràfic de Catalunya, Barcelona. 597 p. ISBN: 978-84-412-2249-6. https://www.termcat.cat/ca/Diccionaris_En_Linia/197

Xue, J., Su, B. (2017) Significant Remote Sensing Vegetation Indices: A Review of Developments and Applications. Journal of Sensors, ArticleID 1353691, 17 pages. https://doi.org/10.1155/2017/1353691

Dialog box of the application

Indexs dialog box