Thoughts on PDF sets and SM input:

-----------------------------------------------------------
1. Make use of the new PDF4LHC15 sets -
does not affect Higgs masses and mixing if we stick to the old SM input:

PDF4LHC15_nnlo_mc was identified to be not a good set for bbh 5FS.
A corrected set provided by Marco Bonvini, which has a fixed
bottom mass threshold at mb=4.58 can be found here:
http://desy.de/~sliebler/PDF4LHC15_nnlo_mc__mb4.58_mub4.58.tar.gz
For ggh these NNLO sets should give results equal to the
original PDF4LHC15_nnlo_mc sets at the permil level.
Since for now we only have the corrected NNLO set,
one can call SusHi with this NNLO set together with the standard
PDF4LHC15_nlo_mc set at NLO. The NNLO set is then employed for bbh,
and for ggh where both NLO and NNLO set are of relevance we rely
on the fact that the differences between PDF4LHC15_nnlo_mc
and PDF4LHC15_nnlo_mc__mb4.58_mub4.58 are tiny.

Similarly Michael can provide his bbh 4FS numbers using the
new original PDF4LHC15_nlo_mc set.

Comment by Michael:
The bbh subgroup is working on an official recommendation for the
combination of 4FS and 5FS, which might also be available as a function
of the Higgs mass?! -> Rather than using our independent calculation
we might just stick to their numbers in the future, which means we
just reweight their bbh predictions. We should discuss this, once
YR4 is public.

Previously the PDF+alpha_s error for ggh and bbh was separately calculated with
MSTW2008 as a grid containing relative uncertainties as a
function of the Higgs mass. We just do the same with the new
PDF sets and provide PDF and alpha_s error separately. Moreover for
bbh 5FS we can also give a mb and mub error, as recently done for
the bbh subgroup.

-> First step: Redo the PDF and alpha_s relative uncertainties
using the new PDF sets, which is mostly independent on the
new SM input. Stefan can provide new files soon.
The stepsize for the SM Higgs mass in those files is:
70-200 GeV in steps of 1 GeV
200-320 GeV in steps of 5 GeV
320-370 GeV in steps of 1 GeV
370-3000 GeV in steps of 5 GeV
For now we provide PDF and alpha_s uncertainties independently.
They can be combined in quadrature to a PDF+alpha_s uncertainty
by the scripts.

-----------------------------------------------------------
2. Adjust the SM input parameters to the newest recommendation:

The relevant parameters are mb(mb), mb^os, mc(mc) and alpha_s
(mt^os is the same as before).
Moreover the lepton masses are of relevance for FH and BR calculations.
alpha_s for XS on the other hand is taken from the PDF sets directly.
--> If we make use of the new SM input we obtain a new Higgs mass from FH,
which in turn means we need also new BR files with the same
input leading to the same masses, mixing, etc.
--> The BR files have to be redone!

Comment by Michael:
The SM input is in turn dependent on the model under consideration.
Is there an estimation to which extend the SM input is affected?
We should keep this in mind.

-----------------------------------------------------------
---> To do:
1. Stefan will redo the files containing the PDF and alpha_s
uncertainties for ggh and bbh 5FS 
and also add mb and mub uncertainties for bbh 5FS.
Michael needs to provide his 4FS bbh numbers calculated
with the new input and PDF set.

2. Sven needs to redo all the BR files.

3. We can produce new XS for 13 and 14TeV (including scale
uncertainties) using the new SM input and the PDF prescription
from above.
Then finally we need someone to work on the combination of
the files. Thanks again to Felix for all his efforts!
All the best!

A crucial issue is how to store the old files. We have to
make sure that everything is well documented!