I want to do water quality analysis using Landsat images. Can anyone help show me how to estimate Chl-a from the Landsat images? In-situ data is required for this analysis or not?
here link presentation file for estimating chl-a
http://glei.nrri.umn.edu/eagle/Documents/Emtgnotes/03meeting/Scaling_Han.pdf
https://www.researchgate.net/publication/255597274_Estimating_and_mapping_chlorophyll-_a_concentration_in_Pensacola_Bay_Florida_using_Landsat_ETM_data
Article Estimating and mapping chlorophyll-a concentration in Pensac...
Dear Jadav,
You need in situ data for the best result. Go to the field and collect your sample with same accuasion time for landsat satellite. Secondly, get the all vegetation indices like NDVI, NDWI, SAVI, GVI and more. Also you can use the some band combination like 4+5+6 especially vegetation bands is important. Firstly apply the parametric models like stepwise regression and you may try the non-parametric ones such as regression tree, artificial neuron network and more.
Chlorophyll-a absorbs most energy from wavelengths of violet-blue and orange-red light but it reflects green/yellow light, and as such contributes to the observed green color of most plants and we know that plants reflect mostly in the infra-Red, you can estimate chlorophyll-a using the vegetation index or NDVI from the Landsat images and make use of Water Index (NDWI) using band green (water gives a high reflection in this band) and the red band (water reflection really drops in this band).
Thank you. I have tried the bands 3(green) to 5 (NIR), but I must wait to the next season for in situ measurements. I can see that the regressions from elswhere give all too high chlorophyll values in my area. Is this due to latitude dfferences or is it due to contents in the water (everything unless Chl)? What about different satellite images over neighboring areas. Must the algoritms be adapted to each image?
Arne
Hello everyone, Jodom here.
Is there any way to estimate chlorophyll-a from Landsat imagery without having to use in-situ datasets ?
I'm posting a further explainer below:
Challenge:
I'm looking to measure water quality parameters from Landsat imagery. As pointed out in the thread above, I need in-situ measurements for correlation and then developing a model for the measurements.
Given I am doing analysis using historical data, I possibly can't get in-situ datasets for the time period and epochs (1980 - 2016, 4 year epochs, Lake Victoria), is there any other way of estimating the water quality parameters(any parameter, chlorophyll-a most recommended) without having to use in-situ datasets and correlation?
Thanks
@Konuko, there are several globally developed Landsat chl-a algorithms found in literature that you can use to estimate chlorophyll. Fin attached one document I recommend you to read Article Using Landsat to extend the historical record of lacustrine ...
Hello Collins Acheampong, thank you for your recommendation. Let me get the full text and review the procedure.
Regards
Thank you all! Just like Konuko, I want to know if I can use some measure of change instead of an absolute value, analogous to absolute/relative abundance parameters. Moreover, I would like to know if there are R packages developed for that. Thanks!
Ulises
kindly help me and tell the steps to estimate chlorophyll in River Water from Landsat 8 imagery......
Thanks....
Konuko Jodom , do you have updates regarding your study? I'm currently studying the water quality of a river and I am looking for the best algorithm to use, without taking in-situ measurements. You might have some recommendations.
Hi I am working on lake water quality. I am gathering both in- situ and Landsat 8 data. Can anybody help me read chlorophyll value from Landsat 8 images?
- Badges
- Science topic
- Similar topics
- Geoscience
- Remote Sensing