I am working on getting time series NDVI data from Sentinel-2 L1C, since it is available before Sentinel-2 L2A. Is it ok to use L1C product or I need to use L2A to develop time series NDVI?
NDVI will do some normalization that can help in reducing the atmospheric effects. If it is relevant to your project this is another question.
You can check it, up to a level....
You can use old imagery compare the NDVI values for the two products.
There might be a paper in it.... compare the NDVI by the products for different landscapes and atmospheric conditions.
In general, from the physical point of view, you always have to use Level-2 because you are measuring surface reflectance using surface reflectance product this way. In reality, in summer conditions the difference between NDVI is not to large, but it is just not correct to use Level-1 for vegetation monitoring.
Another suggestion is using some more modern indices which take into account the variety of wavelengths available to Sentinel-2 rather than outdated NDVI, which is not very accurate and inclined to early saturation.
In addition to previous answers, is worth spending some time assessing which atmospheric correction algorithm you can use. Some time ago, sen2cor (the main algorithm to generate Level-2 products) didn't take into account atmospheric visibility in rural areas.
- Badges
- Science topic
- Similar topics
- Geoscience
- Remote Sensing