Explanation of notes in table:

* = Confirmed leader of investigation team
*(H,L) - Confirmed leader of (Hanford, Livingston) part of investigation
S = Special conditions may be required at times during data run, perhaps dedicated shift
P = Special preparation may be needed before start of E5 run

• Quantify correlations between GW channel and other channels.

 Such as laser frequency or intensity. Investigate use of cross-correlations to reduce noise in GW channel.
• Quantify correlated ambient noise between sites.

Identify ambient noise sources correlatated between the sites, including artifacts due to 60 Hz synchronization and common data acquisition electronics, including computers.
• Quantify correlated environmental transients between sites (including mains)

One hopes for minimal environmental correlations betweens sites, but correlated transients in electrical mains (60 Hz & harmonics) may be seen. Magnetometers and seismometers should also be checked
• Identify & catalog environmental disturbances

(includes impulses, bursts, chirps, contributions to non-Gaussianity).
A large variety of transients from the environment have been seen in previous data taking runs, some but not all of whose sources have been identified. Although this investigation would naturally focus on transients seen in the main GW channel, other channels to examine include those in the PSL and Mode Cleaner. It is also desirable to quantify the contributions of transients to non-Gaussianity in the GW channel.
• Quantify calibration stability & stationarity of data.

The calibration transfer function of the GW signal may vary in time, with laser power, with gain in the servos, and from other factors. Characterizing how stable the transfer function is will help determine how frequently different calibrations will need to be performed. The interferometer's sensitivity to gravitational waves also depends on its noise level, which may not be stationary, being affected by the factors above and by many others. The dependence of the interferometer noise in different bandwidths on temperature, lock-point of MC, etc., should be checked.
• Investigate angular fluctuations.

Quantify relative angular fluctuations between the local frames of different buildings. Look into the consequences of large angular deviations on stability of servo systems and sensitivity levels.
• Check data against detailed tidal prediction.

At Hanford, verify that residual tidal variations after tidal actuation are small. At Livingston verify that the tidal model for a single arm agrees in magnitude and phase with predictions.
• Investigate sources of lock losses.

This is open-ended, depending on what is seen. At Hanford one expects interferometer instability and transients to play a role, while at Livingston limited dynamic range is likely to lead to periodic lock loss from tidal drift.
• Quantify timing precision (intra- and inter-site).

Include the entire interferometer chain locally - from mirror excitation/motion to acquisition system readout. In E2 small anomalies were seen. Verify that a timing signal injected simultaneously at both sites is seen as such downstream (e.g., apply a separate GPS clock trigger to a data channel at both sites).
• Check data integrity end-to-end.

Make sure data is valid when it is read out, written to disk, transferred to tape, sent to the CIT, injected into the archive, and read back from the archive.
• Check data merging.

 Verify that an analysis using corresponding channels from the two sites sees sensible results (e.g., correct channel names, sampling rates, calibration scale factors).
• Quantify strength and stability of line noise in GW channel.

60 Hz and its harmonics are expected to be apparent in the GW channel. Study amplitude and phase stability of the contamination through course of run. Look for correlations between strength and IFO or environment conditions. Characterize violin modes and test mass vibration modes, i.e., Rayleigh behavior consistency, and correlation with IFO or environment. For example, is there ringdown immediately after lock acquisition? Is there growth?
• Test simulated astrophysical signal injection

Although the four upper-limits groups plan elaborate simulation efforts to verify analysis code, it is also desirable to verify the entire data pipeline from test mass motion to signal detection. Hardware and software are now available to allow injection of simulated gravitational wave forms by stimulation of test mass coils. It is important to verify in advance of the E7 run that this signal injection infrastructure works properly.