$Header: /nfs/slac/g/glast/ground/cvs/irfs/caldb/CALDB/data/glast/lat/bcf/ea/README,v 1.1 2004/08/04 18:35:10 jchiang Exp $

The file aeff_DC1.fits contains the LAT effective area as a function
of incident photon energy and incident angle (theta) measured with
respect to the LAT z-axis.  Azimuthal (phi) symmetry has been assumed.

The file has three binary table extensions:

EA_FRONT      effective area for front-converting events (HDU=2)
EA_BACK       effective area for back-converting events (HDU=3)
EA_ALL        effective area for all events so that 
              EA_ALL = EA_FRONT + EA_BACK (HDU=4)

Each extension comprises a single row that contains the following
variable length vectors:

ENERGY_LO      lower bound of the 36 energy bins (MeV)
ENERGY_HI      upper bound of the 36 energy bins (MeV)
THETA_LO       lower bound of the 20 theta bins (degrees)
THETA_HI       upper bound of the 20 theta bins (degrees)
EFFAREA        720 values of effective area for each bin (cm^2)

For the EFFAREA vector, the theta bins run faster than the energy
bins.  IDL code for accessing the bin corresponding to an energy E and
angle theta would look something like

EFFAREA(max(where(E gt ENERGY_HI))*20 + max(where(theta gt THETA_HI)))

Note that for some reason THETA_LO[0] in the FITS table has a small,
non-zero value.  This is a flaw in the way the FITS files were
generated from the original ROOT file. Its value should be zero.