If you are using a laser guide star as your adaptive optics reference then it is useful to know how the sodium layer changes during the night and during the year. ALFA and NUI Galway (Ireland) have worked together to design and build a system to monitor the sodium layer very easily. It works on the LIDAR principle, in which we measure the time of flight of photons from the laser to the sodium layer and back to the AO bench. This can give extremely accurate information about the height, profile, and column density of the sodium in a very short time.
It is a novel system to implement on ALFA because we have a continuous wave laser - and so this must be modulated. We send up a random sequence of microsecond pulses, and measure the returned flux every 0.25 microsecs. The information we want is recovered by cross-correlating this data with the original pulse sequence. The total time to obtain a measurement and recover the sodium profile is less than 1 minute. With reasonable atmospheric conditions we will be able to obtain very high quality data with a height resolution of 150m.
| 17 October 1999 | 18 October 1999 | |
|
|
|
|
| Best results can be achieved if JavaScript is enabled; otherwise click on the images to see an animated version | ||
Above is the very first data we acquired, during 2 nights in October 1999, showing 3 8-min sequences of 30-sec integrations spread over an hour or so on each night (separated by blank frames). The data have been smoothed with a 500m digital filter since the signal was rather weak, due to very bad seeing and thin clouds. As a result of this we also could not calibrate the sodium column density. But the height, around 90 km above Calar Alto, and the profiles are not affected by this.
|
click on these buttons for images showing the variation in either the
peak or the centroid height in the sodium profile on each night
|
|
height of the peak of the profile. On the first night, this traces the position of the prominent feature seen on the left-hand side above, which has a very clear movement by more than a kilometre on a timescale of 10 minutes. A bimodal distribution is apparent on the second night, due to slight changes in the relative strengths of the two peaks seen on the right-hand side above. The separation of these peaks stays roughly constant while they both descend nearly 2 km through the layer during one hour.
centroid height of the profile. This has much more important consequences for wavefront sensing, since any change in centroid height will be detected as a defocus and applied to the science observations. On each of these two nights it was fairly constant, but differed by about 1.5 km between them. The changes of up to 500 m which are apparent do not matter for K-band observations with ALFA on the 3.5-m telescope. But it would become a problem for J-band observations on an 8-m.
Eventually we plan to have a system that allows us to make a quick measurement of the sodium layer at any time.
Many thanks to all those whose help has been invaluable to the success of this project