Added a jupyter notebook for converting from radiance to reflectance

[MattFerraro]

Mostly this is a copy of the NDVI notebook, with some more text focusing on the Radiance aspect

[MattFerraro]

You can view the notebook here: https://github.com/planetlabs/notebooks/blob/toar/toar/toar_planetscope.ipynb

[NickDub2]

Matt this is great! will give it a look now

[NickDub2]

Might want to use 'spectral band' rather than color band here, with analytic products we tend to avoid the use of the word color

[MattFerraro]

Agreed, done

[NickDub2]

same comment as before around using spectral band

[danabauer]

Again, you could skip the download step and refer users to the NDVI guide. Your guide could start with "extract data from the spectral bands" or it could start with Extract the coefficients from the metadata.

[MattFerraro]

That's true Dana, but I really wanted this guide to stand completely on its own as a usable product in the same way the NDVI one does.

If we really wanted to break things down we could have one guide that just does downloading, then several others that all work on that same downloaded image, but I think it would interrupt the flow and self contain-edness of the notebooks

[NickDub2]

I think this step is "Save the Reflectance Image". Not sure if you want to include this here, it is often the case that reflectance values are scaled by factor of 10000. This allows the image to be saved as an unsigned 16bit-Int, the same format as the native scaled radiance

[danabauer]

^^ +1 to @NickWIlson33's suggestion. That's useful info to include.

[MattFerraro]

Agreed! Changed.

[NickDub2]

Reflectance Values rather than Radiance Values

[MattFerraro]

👍

[NickDub2]

same nit: reflectance rather than radiance

[MattFerraro]

thanks for catching all these!

[NickDub2]

nit: reflectance rather than radiance

[MattFerraro]

🥇 for attentive reading!

[NickDub2]

Couple small things, but LGTM.

[Chandler]

W/m^2 -> W/m^2/sr

[kjordahl]

even better, W/m$^2$/sr to get a real superscript

[danabauer]

Using LaTeX in Jupyter notebooks: http://data-blog.udacity.com/posts/2016/10/latex-primer/

[MattFerraro]

fixed!

[kjordahl]

True, but reflectance is also dependent on spectral response of the filters.

[MattFerraro]

I think the next paragraph does a good job of addressing that

[danabauer]

This is great, Matt! Just a few suggestions. Excited to get this up.

[danabauer]

Using LaTeX in Jupyter notebooks: http://data-blog.udacity.com/posts/2016/10/latex-primer/

[danabauer]

You could skip listing the steps and requirements and tell users to follow the NDVI guide through the few steps. This guide should focus on the conversion from radiance to reflectance.

[danabauer]

Again, you could skip the download step and refer users to the NDVI guide. Your guide could start with "extract data from the spectral bands" or it could start with Extract the coefficients from the metadata.

[danabauer]

^^ +1 to @NickWIlson33's suggestion. That's useful info to include.

[danabauer]

This is a big block of text at the top of a notebook. Perhaps break it up with a new topic heading or drop in a diagram/pic of a satellite showing where the spectral filter is

[MattFerraro]

Agreed, I broke it up with a little image

[nateweisiger]

@MattFerraro - Can you rebase this branch with master, and put your notebook in the new "jupyter-notebooks" directory? Master was rearranged a bit since opening this PR.

Thanks!