Can anyone help me to define a range of NDVI value to extract bare soil pixels for Landsat TM?

Hi Dat,

In fact it depends on your image ! So on your study area !

The general rule is to test different threshold values and check visually on your image.

A.Safia

you can select range of 0-0.05 as bare soil....

It is better check ground truth points and verify the values for bare soil. You might select several non-vegetation areas with the bare soil.

Thanks for your comments. I have checked website of NASA. It indicated that very low values of NDVI (0.1 and below) correspond to barren areas of rock, sand, or snow. Moderate values represent shrub and grassland (0.2 to 0.3), while high values indicate temperate and tropical rainforests (0.6 to 0.8). 

Here we would like to extract bare soil pixel from Landsat imagery based on NDVI value fro semi-arid and arid regions. Any comments would be highly appreciated.

Have a nice weekend

Hi Tien Dat Pham,

I agree with previous comments, that it depends on your image and characteristics of your target area. It might also be worth considering other indices like for example the SAVI (Soil-Adjusted Vegetation Index) (visit:http://www.seos-project.eu/modules/agriculture/agriculture-c01-s04.html for general introduction) or one of many other possible indices (you might take a look at this index database:http://www.indexdatabase.de/)

Good luck with your work!

Hi dear, Tien Dat Pham

you can follow 

one or less                          ---  barren rock, sand or snow 

approximately 0.2 to 0.5   ---- shrubs or agriculture

approximately 0.6 to 0.9   ---Dense vegetation (Forest) 

HI Mr. Tien Dat Pham,

I agree with the former comments that the the NDVI could varies depends on different type of satellite images, season, study areas, atmospheric effects, soil type, and humidity, and so on. Therefore, to know the accurate range of NDVI for your case study areas, you should have some in situ or other kind of reference data to analyze and derive the most suitable threshold for your study site. The common value could be around 0.1 or below (as you said); however, NDVI <= 0 or -0.5 normally represent water body or wetland.

Wish you the best in your academic research.

Hi all

I am studying NDVI for my research.

I get negative values (-0.1 to -0.22) for bare soil. I used google image as a reference. Is it ok to have negative values for bare soil in NDVI?

Thanks

Dear Choman Salih;

I think that negative values of NDVI may come from wet soil. Please check the data acquisition date. Maybe it had a rain before the date you used satellite imagery.

Basically, negative values of NDVI represent water bodies, bare soil is normal positive values

Cheers

Thanks for your answer Mr Tien Dat

Cheers

NDVI Threshold value to distinguish between bare soil and build-up?

its depend on your data. generally, NDVI value of 0 or near 0 shows bare soil whereas towards the -1 shows water-related features and towards +1 shows vegetation coverage

it it less than 0.1

Dry Bare soil is 0.025

http://www.indexdatabase.de

In fact it depends on your image and your study area .

Basically, in my reaserach the negative values of NDVI represent water bodies also bare soil negative values

It depends on the image and study area.

Usually less than 0.2

Hi,

It depends on the study area, but it is usually between 0 and 0.1.

b1 ge 0 and b1 lt 1 or

b1 ge 0.14 and b1 lt 0.18