10.10 Failed Observations

HST observations fail at a rate of a few percent. Some of these failures result from occasional guide stars that cannot be acquired, or from an instrument anomaly, or the telescope happening to be in a safe mode when a particular observation was scheduled. Such failures, which are obviously beyond the proposer's control, can usually be scheduled for a repeat observation. When this is the case, the proposer receives a notice of the failure and information on obtaining a repeat observation.

A smaller fraction of failures do not have a clear cause, and may not be evident from our internal reviews of data quality. If you believe your observation has failed or is seriously degraded, then you may request a repeat using the Hubble Observation Problem Report (HOPR) Web Form. The HOPR must be filed within 90 days after the observations are taken. In cases where the failure resulted from proposer error (e.g., incorrect target coordinates), a repeat will not be granted. In cases where the failure was a result of incorrect instrument performance, or incorrect information provided by STScI, a repeat is usually granted.

The policies that apply to failures and repeats are described in the Policy Document for the Telescope Time Review Board (TTRB), available on the Web. We wish to emphasize in particular:

• Standard policy dictates that if observations are to be repeated, the degraded/failed observations will be made public.
• If an observer has obtained more than 90% of the planned observations and the missing data are not uniquely important, then a repeat is not normally granted.
• If a snapshot or pure parallel exposure fails during execution it will not be repeated, regardless of the cause of the failure.
• Observations taken using Available-but-Unsupported modes that fail due to the use of the unsupported mode will not be repeated.
• Observations that are lost due to bright object violations will not be repeated.