If Corresponding bands of 3 and 4 in Landsat 7 are bands number of 4 and 5 in Landsat 8, So to calculate NDVI by Landsat 8 imagery we should use:
NDVI = (NIR − red) / (NIR + red) = (Band5 - Band4)/ (Band 5 + Band 4)
Do anybody agree? If yes so my other related question is:
There are some wavelength incoherent between Landsat 7 and 8 bands. Can we connivance mentioned wavelength differences to NDVI calculations?
Yes I agree, 4 and 5... Bear in mind the differences in the frequency band width (range) for red and NIR will be a bit different between the two sensors.
questio in the question ;-)...is it possible to uuse band 6 (swir1) or band7 (swir2)having a better NDVI?
To: Giovanni Pecora
Actually, I tinkered with substituting bands 6 and 7 for NDBI, and liked the result.
Dear everybody
Thank you so much because of your participations in answers
Dear Phillips
At first I appreciate you for your participation again.
secondly, Is it possible to explain your answer to Giovanni Pecora clearly?
Yes,
Substituting bands 6 and 7 proved more effective as NDBI (Normalized Built-up Index) components, then as NDVI components.
example: with TM imagery the NDBI = (MIR - NIR) / (MIR + NIR). Therefore, substituting OL 8 bands 6 and 7 for the MIR component yielded some interesting results - band 6 produced more defined impervious identification than did band 7 - but still worthy of further investigation since I have only looked one well known study area.
Using these bands as substitutions for the MIR component of the NDVI, however, yielded some confusing results - vegetation was not as clearly defined as I am accustomed in a TM NDVI calculation.
That said: for any one who can answer Saeed Najafi's related question, in the initial post, I would be very interested in the answer.
.