add more realistic METimage RSRs

[rolle-at-work]

Updated to more realistic, though not measured spectral response functions for METimage

• Closes #xxxx
• Tests added
• Tests passed: Passes pytest pyspectral
• Passes flake8 pyspectral
• Fully documented

[coveralls]

Coverage remained the same at 71.113% when pulling 1cf912c on rolle-at-work:add_simulated_metimage_rsr into 0f6f145 on pytroll:master.

[rolle-at-work]

The new file looks good and i tested reading with hdfview and a pyspectral plotting routine.

[adybbroe]

@rolle-at-work I think you could/should use the todf5 function in utils.py. Then we will also get rid of the redundant det-1 layer. There is only one detector, and in that case I think we should not use that extra layer

[rolle-at-work]

@adybbroe But is this layer also gone for AVHRR and other one-detector instruments? I think this layer was usefull to allow for a general reader of bof both single and multi detector instruments.

[adybbroe]

Yes @rolle-at-work they are gone for all single detector sensors. They should be at least. I just checked a couple. And the reader can handle both, no problem

[adybbroe]

Oh well, now I look back in the history here:
#68
Seems like we were preparing for more detectors. So, if we know there will be, and I gues we do know that @rolle-at-work? Then we should keep it as is I think. Sorry

[rolle-at-work]

Yes @adybbroe METimage will have 24 detectors per channel

[adybbroe]

@rolle-at-work I have uploaded an updated RSR data archive with your new Metimage data included, please check it out and adapt your PR accordingly. You will need to change version and url of the RSR data in utils.py Version should now be RSR_DATA_VERSION = "v1.0.15" and the url is now https://zenodo.org/record/3824535/files/pyspectral_rsr_data.tgz

[adybbroe]

So, @rolle-at-work, could you please update utils.py according to my comments

[adybbroe]

LGTM