Observing began in semester 2003B. By the end of 2008, every survey field had been observed at least once. The survey will be formally complete once it is confirmed that acceptable data quality has been achieved in all fields.
This is all Galactic longitudes in the Northern Plane within the latitude range -5° < b < 5°, a total of 1800 square degrees. The INT Wide Field Camera (WFC) offers a pixel scale of 0.33 arcsec per pix, allowing on-sky structures above 1 arcsec in size to be resolved in principle.
At each pointing, sequences of narrow-band Hα, Sloan r' and i' images are obtained with the INT's Wide Field Camera. The Hα filter has a bandwidth of 95 Å and is centred at 6568 Å. For the accompanying broad-band filters we have preferred the Sloan filters because of the more nearly rectangular shape of their pass-bands. The aim is to cover the survey footprint twice, observing once at each of a set of specified field centres that combine to cover the plane once, and then again at a second set of offset positions. This strategy enhances our quality control and ensures that strips of sky, otherwise falling in the gaps between CCDs, are not left out altogether. The total number of field centres demanded by this strategy works out at 7635 x 2. The telescope time needed to take data at each field position (plus its offset) and move to the next field centre is approximately 10 minutes.
The faint-end limiting broadband magnitude deemed acceptable for release is r=20 or better (5σ). In most data, it is rather better than this (r = 21-22). Hα filter exposure times were fixed at 120 secs throughout as a compromise between approaching the broad-band magnitude limit (it is typically a magnitude brighter), and ensuring the total survey execution time was feasible. The narrowband sensitivity is similar to that of the AAO UKST H&alpha survey. The i band faint limit is also typically about a magnitude brighter than in r. Since most fainter Galactic Plane objects have typical r-i of ~1, the number counts in r band and i band are not very different. In median seeing, saturation is around 13th magnitude.
Much of the data-taking was performed in bright time, except when the bright moon falls in the Galactic Plane (times at which grey/dark time was essential). For those fields observed near times of full moon, this has resulted in considerable variation of background levels in the images over and above that due to diffuse Hα, which need to be addressed before the data can be mosaicked successfully.
The target seeing limit for data to be counted as meeting survey requirements is 2 arcsec. Some data at worse seeing in lower density fields when the sky transparency was nevertheless good will be released. Median seeing in survey-standard data is 1.1 arcsec. Standard star observations are typically obtained every 2-3 hours to facilitate photometric calibration of the survey data. The external, or systematic, errors are expected to be in the region of 0.03 magnitudes. We calibrate in the Vega system (not AB).