The large data samples obtained will be analyzed using information from the FPD in conjunction with information from other DØ sub-detectors as discussed below.
of exposure
with a 25% efficiency, 25% acceptance (averaged over p's and 
's),
and 10 events needed for discovery. With a single
interaction requirement the minimum observable cross section
will be about 0.5 pb.
For the double pomeron case, the relevant acceptance is about 0.5%
resulting in a 50 times higher minimum cross section of about 25 pb.
For those processes that we observe, we will
measure various kinematic properties
such as 
and |t| using the pots, and
jet 
and 
, energy flow, etc., using the rest of the DØ detector.
distributions of jets (shown earlier) or
electrons from W boson decay, both of which are
sensitive to the hardness of the partons in the pomeron.
Another variable, used by the UA8 Collaboration with only 100
dijet events, is the longitudinal momentum of the two-jet system
in the diffractive center of mass [5]. This variable
directly reflects the imbalance between the parton in the pomeron and
the parton in the proton, and, in combination with known proton PDF's, allows
us to measure the pomeron structure function. We will also be able
to input HERA pomeron structure functions obtained from deep-inelastic
scattering and test whether they describe our data.
and |t|
bins. One way to express the results is to take ratios of hard diffraction
and inclusive diffraction, giving quantities such as the fraction of
diffractive events in a given and |t| bin that have jets,
W bosons, top, etc.
Each ratio can be directly compared to predictions for
different pomeron structures and different quark and gluon fractions.
Current Monte Carlo's assume no |t| dependence
of the results. We will be able to test the validity of this assumption
simply by determining this ratio for a few |t| bins (for a given process
and range). This approach removes acceptance errors including the
error from extrapolating to |t|=0.
We will also be able to make absolute cross section measurements.
This requires accurate knowledge of our acceptance, and
is where the full power of the FPD is demonstrated.
By using symmetric pots (up-down or left-right), we can
determine the beam position at each quadrupole station
(Sec. 5.1) to an accuracy of about 100 
m.
Averaging the cross sections measured in symmetric spectrometers
reduces the cross section error from the beam position to a
second order correction (which contributes less than 5% to the
total acceptance uncertainty).
We can then use well-measured tracks in the quadrupole spectrometers
to calibrate the dipole spectrometer (for tracks with |t|>0.5 GeV
we have a large overlap between anti-protons in the two types
of spectrometers). Conversely, we can use the |t| distribution measured
in the dipole spectrometer (which has full |t| acceptance for
) to extrapolate the quadrupole measurements to |t|=0.
The dominant error in the acceptance is due to the dispersion of the beam,
which causes about a 15% uncertainty integrating
over the |t| range of the quadrupole spectrometers.
We can reduce this error significantly by measuring the dispersion using
elastics and then unsmearing the |t| distribution. We expect final errors
on the cross section due to acceptance
of a few percent in the dipole spectrometer and less than 10% in the
quadrupole spectrometers.
For a fixed |t| bin, the error in the diffractive mass is 
(6% for 
, 3% for 
, etc.). Therefore,
from the FPD point of view we will be able to make accurate
cross section measurements.
For jet
cross sections we will be limited by energy scale errors which will
likely be 30-50% for our range. Using ratios will
allow us to reduce these errors.
This time we will
measure the fraction of jet events above an threshold that
are diffractive (similar to the rapidity gap measurements), making
us insensitive to energy scale errors. We thus expect to
measure all cross sections (or ratios of cross sections) with errors
of 10% or less.