I am trying to estimate chlorophyll-a by using Landsat/ MODIS image. So I want to know atmospheric correction for the above analysis is required or not? There is so much technique for atmospheric correction like DOS, COST model, FLASH ATCOR etc.Which is the best accurate method?
As you have found out there is a long and extensive research history on remote sensing (RS) of pigments. There are as many variations to conducting it. In addition to the reference already provided, a great background on RS of leaf pigments is available in papers by Gitelson and others (https://www.researchgate.net/profile/Anatoly_Gitelson/publication/241061369_Remote_estimation_of_canopy_chlorophyll_content_in_crops/links/02e7e5305b51d55ebf000000.pdf)
Gitelson covers a ratio that is directly useful in the image sources you mentioned. we have applied the various relationships he suggested and found the NIR/green ratio works well up to a point and the NIR/red in others.
As for atmospheric correction when applying optical RS data. I suggest correction is necessary if you intend to map subtle changes...i recommend some of the earlier papers by Huete (https://cimss.ssec.wisc.edu/dbs/China2011/Day2/Lectures/MODIS_MOD13_NDVI_referenc.pdf)
(http://research.eeescience.utoledo.edu/lees/papers_pdf/Huete_1997_RemSensEnv.pdf)
as for techniques to correct the atmospheric component (scatter and attenuation, and etc) I also recommend the simply approach of using a low reflectance target to remove bias (dark pixel subtraction is sufficient in many cases, see Ahern and others for background)...if these are not available, and no quasi constant reflectance targets exist in the scenes, then other approaches have to be used...there are many,,,you might look at Hall's papers and those that started with his approach...and combine that with your ATOR or other transfer methods...you can obtain estimates of the optical depths online and use these a s starting estimate to the transfer calculations (ATCOR and others require this estimate or a variation of this measure as the horizontal visibility)....
hope your project is successful
Article Remote estimation of canopy chlorophyll in crops
Your challenge is the low spectral resolution of landsat. However, there are already studies about it, thus, you now have more references. One study posted below applies atmospheric correction. (By the way, applying atmospheric correction is not very hard to do.)
I also attached a study that posed the question of "when and how to apply atmospheric correction."
Cheers!
http://ieeexplore.ieee.org/stamp/stamp.jsp?tp=&arnumber=5278182
http://www.unc.edu/courses/2007spring/geog/577/001/www/Song01_RSE.pdf
- Badges
- Science topic
- Similar topics
- Geoscience
- Geoinformatics