H2 spectrometry?

[Tony Bryant]

My objective is to build an A/F meter that is accurate between 11-
14 A/F ratio. Ion pumped oxy sensors (besides costing too much), 
seem to be good at measuring in the lean area, but not rich (as 
there is precious little variation in O2 concentration in 14 & below 
ratios). 

Looking through Heywood, I've come across a few interesting facts:

1) O2 concentrations are essentially zero, when rich, rising 
exponentially when going lean

2) CO concentrations are the opposite, zero when lean, rising 
exponentially when going rich.

But importantly:

3) H2 concentrations are about 1/2 CO concentrations, at all 
mixtures.

CO can be detected using IR spectrometry, but H20 has an 
overlapping range, so exhaust gas must be dried to get an 
accurate reading. This is obviously impractical for a view as you 
drive type guage.  

But the question is, is H2 spectrometry doable in this case?