NDVI from Surface Reflectance and TOAR products
I've been working with data from both products (SR and TOAR), and I realized TOAR gives better results in distinguishing vegetation/no vegetation in my area of study. However, I'm a little confused since apparently, SR is an improved version of TOAR products? Or perhaps my understanding is not pretty clear about it. For example, I use 0.2 as a threshold, in order to get rid of vegetation and as a result, TOAR products give me a consistent output over time whereas SR is sometimes recognizing vegetation, and sometimes not.
Thanks for helping me with this!