The first step in planning this observation is to decide on the instrument and its configuration. There are two imaging cameras on HST that we could use, WFPC2 (Wide-Field Planetary Camera 2) and FOC (Faint Object Camera). You might think that from the name we would pick WFPC2 but that's not the case. There is one fundamental tradeoff between the two cameras that makes the decision easy. The pixels that make up an image with each camera are of very different sizes between WFPC2 and FOC. The FOC has smaller pixels than WFPC2 and will capture the maximum resolution image possible with HST. The WFPC2 was built to take picture over a wider area of "sky" and had to sacrifice resolution to get more area.
Having decided to use the FOC, the next step is to decide on which filters must be used. These filters select a limited range of color somewhere between the ultraviolet and visible parts of the spectrum. We have time in a single orbit to take a picture in two colors. There is a long list of possibilities that I won't list here but I've decided on F410M and F278M. These "codes" indicate the type and color of filter. M stands for medium and refers to the width of the filter in wavelength (other possibilities are N for narrow and W for wide). The number refers to the wavelength of light near the center of the filter. 410 means 410 nanometers or 4100 Angstroms and corresponds to what we'd call blue light. 278 nm or 2780 Angstroms is ultraviolet light that we can't see with our eyes.
I've chosen these to be the same as the filters used in the previous observations. Why did we chose these originally? Well, we need to take one picture at a wavelength that has been thoroughly studied in the past. A great deal of work has been done on the appearance of Pluto in blue light, some of which I've done myself. So, the choice of F410M was an attempt to get a picture that can be directly compared to previous work. The choice of F278M was a compromise. Going further and further to the far UV becomes more interesting because Pluto might look dramatically different. However, our sun doesn't actually put out that much light in the ultraviolet so the amount reflected from Pluto drops dramatically as we go to shorter and shorter wavelength. F278M is about as short in wavelength as we can go and still get a decent picture in the time we have.
So, are we done? Well, not quite. We've chosen the filter, now we need to decide how long to integrate in each filter. The FOC is carefully designed to take images of faint objects. You might think that Pluto is a faint object but in fact it's quite bright. Before you can take a picture you must first calculate exactly how bright Pluto will be as seen by the camera. The FOC can see faint stuff but it takes pictures by literally counting the photons as then come in. For an object like Pluto, we must ensure that the camera doesn't have to count any faster than about 1 photon/second in the brightest pixel on Pluto. That's pretty slow. Using the previous observations, I know that the count rate will be about 8-10 counts/second. To keep from damaging the instrument, we must put in a neutral density filter (sort of like using sunglasses) to make Pluto appear dimmer.
With this slow count rate, we will need to integrate on Pluto as long as we can. We have 1 orbit to do this. An orbit is nominally 94 minutes long but during half of this the Earth is in the way and we can't see Pluto. To make the observation the telescope is moved to point at Pluto. As soon as the earth is out of the way it begins looking for guide stars to lock onto and keep the telescope pointed at Pluto. This step takes about 12 minutes. We now have about 40 minutes left before the earth blocks our view again. It turns out that we can get two 15 minute exposures packed into the viewing time and that's what I've put into the schedule.
There is one exciting thing about these pictures. We have the benefit of seeing the previous pictures and using them to modify the experiment. The first time we did this we had to compute the brightness of Pluto without any example to follow. Since there are always some uncertainties involved we had to choose a conservative amount of neutral density to ensure that Pluto wouldn't appear too bright. Well, looking at the numbers yesterday I found that we can use less neutral density than before. This means we will collect 4 times as many photons in these pictures as was done before and thus we should have pictures with less noise.
Now that we know how the observations will be done the next step is to decide when to observe. We've been told by STScI that the observations will be scheduled for the weeks of March 4th or March 11th. During that two week interval, I've calculated when Pluto will show the same side as seen during the other pictures. Those previous pictures were taken at 15, 112, 203, and 289 degrees east longitude. This is a list of when these geometries will repeat.
longitude UT date and time priority ------------------------------------------- 203 1996/03/04 06:11:23 #2 112 1996/03/05 20:56:22 #6 15 1996/03/07 14:14:40 #4 289 1996/03/09 02:51:56 #8 203 1996/03/10 15:29:13 #1 112 1996/03/12 06:14:15 #7 15 1996/03/13 23:32:36 #5 289 1996/03/15 12:09:55 #9 203 1996/03/17 00:47:15 #3
I've chosen longitude 203 as the highest priority because there is an interesting bright spot in the south polar regions and because there are some differences between the visible and UV images. This longitude shows up three times during the two weeks and have been assigned the highest priorities. I suspect that one of these three times can be scheduled but just in case, I've ranked the other times as well. Why do we need a list at all? Why can't we just say when we want the observation to be done?
Well, one of the worst problems plagueing precise scheduling of HST observations is the South Atlantic Anomaly (SAA). This is an area over the South Atlantic Ocean off the coast of South America where the Van Allen Radiation Belts dip closer to the atmosphere due to the shape of the magnetic field of the Earth. If HST is used to take pictures during the passage through this area, the images become contaminated with radiation noise. For any object in the sky, there is roughly 6-7 hours each day where you cannot observe the object because you are in the SAA when the object is not blocked by the earth. Thus there is a 30-40% chance for a random time to be impossible to schedule because of the SAA pasage. By providing a number of possible times, we can be sure that at least one of these can be scheduled free of the SAA.
So, now all this information has been distilled down and transmitted to STScI where the observation plan will be further refined. If all goes well, we just sit back and wait for our data. You might be interested to know that doing all this work and planning took me about 2-3 hours and slightly longer than that to write it all down. Of course, I've got the advantage of having worked at STScI for 3 years learning all this stuff. If you've never worked with HST before, developing an observational plan could take weeks of work.