Fix GSICS calibration in SEVIRI HRIT reader.

[simonrp84]

The SEVIRI readers (NAT + HRIT) have the option to use the default (IMPF) or updated (GSICS) calibration coefficients for transforming digital number into radiance.

The HRIT reader, however, has a bug that means the calibration is invalid if GSICS mode is selected.
This PR fixes the bug by multiplying the GSICS offset by the GSICS slope value, as is recommended in the user guide and as is already done in the NAT reader.

• Tests passed
• Passes flake8 satpy

[coveralls]

Coverage increased (+0.004%) to 90.136% when pulling ac43217 on simonrp84:seviri_gsics_fix into 4aedf55 on pytroll:master.

[codecov]

Codecov Report

Merging #1323 into master will increase coverage by 0.00%.
The diff coverage is 100.00%.

@@           Coverage Diff           @@
##           master    #1323   +/-   ##
=======================================
  Coverage   90.13%   90.13%           
=======================================
  Files         220      220           
  Lines       32510    32512    +2     
=======================================
+ Hits        29303    29305    +2     
  Misses       3207     3207           

Impacted Files	Coverage Δ
satpy/readers/seviri_l1b_hrit.py	92.18% <100.00%> (ø)
satpy/tests/reader_tests/test_seviri_l1b_hrit.py	100.00% <100.00%> (ø)

---

Continue to review full report at Codecov.

Legend - Click here to learn more
Δ = absolute <relative> (impact), ø = not affected, ? = missing data
Powered by Codecov. Last update 4aedf55...ac43217. Read the comment docs.

[mraspaud]

LGTM, I trust you validated with real data?

[simonrp84]

Yes, tested against 4 SEVIRI scenes. Plus, this PR uses the same method used in both the NAT reader and in the EUM ATBD - so we can be pretty confident it's now correct :)

[mraspaud]

So if the same method is used in multiple places, we could refactor the code to use the same function, right? Not saying it should be done in this PR though...

[simonrp84]

I think the readers are so different that it's not worth the effort, tbh.

[mraspaud]

Ok, merging then.

[sfinkens]

Oops, that was me not reading the documentation 🙈 Thanks for fixing this @simonrp84 !