gnuplot> f(x) = c - a*exp(-(x-b)**2/d)

fit [1.55723:1.56540] f(x) "csvch1/F0048CH1.CSV" using 4:5 via a,b,c,d 

Final set of parameters            Asymptotic Standard Error
=======================            ==========================

a               = 0.200041         +/- 0.004786     (2.393%)
b               = 1.56101          +/- 2.408e-05    (0.001542%)
c               = 0.599858         +/- 0.001302     (0.217%)
d               = 9.07425e-07      +/- 3.736e-08    (4.117%)


correlation matrix of the fit parameters:

               a      b      c      d      
a               1.000 
b               0.767  1.000 
c               0.429  0.338  1.000 
d              -0.505 -0.375  0.270  1.000 

sigma = √(d/2)
FWHM = 2.3548 sigma = 0.0015861505252610802
FWHM range [0.0015531552844447974:0.001618473243514764]