Domande e Risposte
Questions collected during the presentation, and coming in via emails - and their answers.
Q. Is the PDF presentation available at the end of the meeting? Thank you very much.
A. Yes, the presentations by Detlef and Tony are available on the website, see here: https://lif.mi.imati.cnr.it/open_search_page.php?query_by=category&search_key=CMN-000000067.
Q. Rotating field (Dobson mount) is a problem for ALFI?
A. No, this is not a problem.
Q. How can I upload my observations?
A. Please register on the website https://lif.mi.imati.cnr.it/ using the 'register' button on the upper right. You will receive a user name and password via email. Log in, and find you will see a menu item 'survey section'. Go there, and follow the instructions.
Q. What is the best gain setting for my camera?
A. Set the gain to see the Earthshine a bit - this helps in the determination of the position. But not much more to avoid the noise.
Q. How accurate should times be recorded?
A. We will use begin and end times to determine the total observation time. This is important even if you did not observe any flashes, to determine flux densities. For the flashes itself, the time is required to be able to check whether a flash was observed by multiple observers. We suggest to record the time to about one second. This is easy to achieve by making sure that your computer clock is synchronised to a time server before the observation.
Q. Do you still recommend the FDS plugin for the FireCapture vídeo recording software?
A. If you have no problems with the Java version, you can use it - just make sure you don't get the illuminated part of the Moon in the field of view. Set everything up in FireCapture, then place a ROI to cover the dark side only.
Q. What's the faintest flash magnitude that could be (has been) recorded with a big telescope against the earthshine?
A. The faintest flash with the 1.2 m NELIOTA telescope was estimated to be ~12th mag in the I filter and 11.5 in R.
Q. If the rate is one every 2.3 hours for brighter than 9 mag. (presentation slide 10), what is the rate for 10 or 11 mag. R?
A. This is shown in one of the plots in the introductionary presentation, showing NELIOTA flashes. NELIOTA recorded a total number of 250 flashes brighter than 10 mag, i.e. on average one every about 1.2 hours. For 11 mag the number is not much higher, as the data is not complete any more.
Q. Is there software out there that is supported by Mac OS? If not, no problem, I would need to fire up my old windows then.
A. ALFI has been developed for Windows, it might be able to work with a Windows emulator. FDS and FireCapture should work on MacOS, but it hasn't been tested by us.
Q. Some videos contain time stamps in the video frames - they might generate lots of false detections, how can we cope with this?
A. In LunarScan, you can actually crop the timestamp out of the selection window. ALFI allows to crop away the border - set it such that the time stamp is outside the detection area. This will mean that you might loose part of the Earthshine area.
Q. Should we switch oj the time code displayed in the video?
A. If the time stamp is near the bottom of the screen you can increase the lower border to mask it out. Or simply switch the time stamp oj.
Q. Are you interested in visual observations - i.e. no recordings?
A. They can be used as independent confirmation of flashes and give the approximate location, so yes, by all means, if all you can or want to do, go ahead with visual observations. Make sure you estimate the time of any flash to within a second. This can be done e.g. by starting to count seconds ("twenty-one, twenty-two...") when seeing a flash, the looking at a clock displaying seconds, and recording the time after ten counts. Then, of course, subtract 10 s to get the time of the impact.
Q. My telescope it too small - only 10 cm - can this be of use?
A. The fainter the detected magnitude, the higher then chances to see a flash. Still, flashes have been observed which are well within the reach of a 10 cm telescope. Thus we welcome observations with any telescope size. Even if you record no flashes - if you provide start and end time of your observations, you can contribute to the determination of flux densities.
Q. I was going to try with my telephoto lens on a camera - will this work?
A. See the previous question - if you are lucky, you will record a bright flash. Having a successful observation with a telephoto lens would be spectacular - but it is rare.
Q. Sharpcap has a feature called "disk stabilization" which will remove a lot of artefacts due to vibrations. Will this work?
A. We haven't tested this. We don't expect this to work for the unilluminated side of the Moon, this functionality was probably intended to keep the bright disks of planets stable (there is a similar feature in FireCapture). But it sure is worth testing.
Q. I can generate only FITS files. Am I out?
A. You can convert FITS to BMP format pretty easily, e.g. using the freeware PIPP.
Q. I normally use an IR-cut filter when imaging. Is it better to remove this if the impact signal is likely to be strong in the IR?
A. We expect LIFs to emit light mostly in the red to near-IR region of the spectrum. If you can, it is better to remove any IR-cut filters.
Q. The graph on slide no. 3 on "Why observing impact flashes?" - why is the X axis in Meteoroid mass? Surely its the KE that matters?
A. The plot is what planetary scientists interests - the flux density as a function of mass. The mass is a direct property of the meteoroid. The kinetic energy depends on the velocity, which changes for particles of the same mass, depending on its orbit and the distance to the Sun. For an impact on the Moon, it also depends on where it hits, as the relative velocity is what counts. Note that the light generated by a meteoroid indeed depends on both mass *and* velocity. However, lab measurements indicate that there seems to be a better relation to the momentum (i.e. mass * velocity) rather than kinetic energy. So, cumulative numbers per area and time as function of mass (or size, assuming some average density) we plot.
Q. LUMIO - Lunar Meteoroids Impact Observer - So the Halo orbit is that orientated along the direction of the Earth-Moon or perpendicular? Sometimes 3D diagrams on 2D graphics can be confusing.
A. The plane of the halo orbit is oriented roughly perpendicular to the Earth - Moon line, but still with some tilt and distortion.
Q. How many impact flashes do we expect? The plot in the presentation on p. 10 says there was about one flash every 5.6 hours brighter than 8 mag. This is what we should be able to see with roughly an 8" or 10" telescope. Will we expect higher rates from the Geminids? Have you looked at NELIOTA Geminid data?
A. The given numbers are just read oj the plot. I.e. NELIOTA has observed one flash brighter than 8 mag on average every 5.6 hours. The limiting magnitude of your system depends on the aperture of your system, the sensitivity of the camera, the exposure time, and the total throughput of your optics. Based on our experience, we estimate that with a 12" telescope you should be able to see 9 mag flashes, i.e. there should be one every 2.3 hours on average. These values are averaged over all NELIOTA data. As stated in the last bullet on p. 9, there will be more events per time during a meteoroid peak. As NELIOTA only observed a few times during the Geminids, it is difficult to provide statistics from that data. This is one of the reasons for this campaign.
Q. What actual science will be or can be done with the data? And for how long should we keep the raw videos?
A. The science will depend on the kind of data we will get. Note that the main goal of this campaign is to motivate people to participate, and to test the technique. For individual events, we can compute the probability of the event being a Geminid, then we can estimate the mass (as we know the velocity; we'll estimate the density). If we have several events, we can compare the flux density with what we would expect from the meteoroid stream models. This requires the people to properly record the duration of their observation, and at least troughly calibrate their magnitude estimates.
Q. I can understand the need for a big aperture to collect more light, but isn't field of view (the amount of the lunar disk covered) also important to maximize out chances of detection?
A. Indeed, that's why the presentation says on p. 15 "As large an area as possible ...". Note that the Geminids only appear on the "upper right part" (with North up) of the lunar disk (the "blue" part in our plots), so that part should be given priority if you have a smaller field of view.
Q. Not everybody has a tracking mount - some just have powerful telephoto lenses - can they take part too just to confirm bright flashes?
A. Observations with any telescope, or even with a telephoto lens, are welcome. The brightest recorded flash was around 3 mag, i.e. it could have been seen with the naked eye. It lasted 8 seconds, so also a long exposure would have captured it. While that flash, recorded by Madiedo and Ortiz (2013), was probably a September Perseid, the Leonids could also create flashes that bright. Chances are low - but if you record one of those, you will have caught a very rare event!
Q. Do we need a dark site to perform these observations?
A. The possibility to detect impact flashes is limited by the background light, mainly coming from the Earthshine on the 'dark' side of the Moon. LIFs have been observed from an balcony of an apartment block near Washington DC. If any bright lights nearby either move the scope to prevent light falling inside the tube or use a chair or large bit of card to block stray light.
Q. In some of the provided sample images, the bright crescent of the Moon is in the FOV. Is this sensible?
A. To maximise the field of view, this is unavoidable. But it depends on the detection software you plan to use. The 'Flash Detection Software' does not behave very well when you have the illuminated part of the Moon in the detection area. It can generate lots of false detections when it seems flickering mountain peaks at the terminator. ALFI is less sensitive to that.
Q. You mention that we should use stars in the Pleiades for magnitude calibration. These are blue - what colour are LIFs? Is it sensible to use the Pleiades for calibration?
A. LIFs are generated by thermal emission of the impact material, i.e. it will be dust that 'glows'. We know, mainly from NELIOTA observations, that the typical flash temperatures are 2000 to 5000 K. That means that the peak of the emission is in the red or near infra-red. The stars we propose as comparison are more blue - they have temperatures around 7000 to 10000 K. Comparing the magnitudes of the two will introduce errors of maybe 0.5 mag. NELIOTA is pointing their telescope to calibration stars at similar elevation above the horizon every 15 min. We think that for this campaign a magnitude estimate of +/-0.5 mag is good enough. This is what we will get if you point your telescope to the proposed set of stars in the Pleiades.
Q. Why is your team concentrating on ALFI - why not Lunarscan or the Flash Detection Software?
A. Lunarscan is only available from the Internet archive; FDS is not behaving well when the illuminated part of the Moon is in the field of view, also it has issues with the latest version of Java.
Q. How much disk space should I allow for on my laptop to video earthshine?
A. Old-fashioned PAL and NTSC were using up 26 GB per hour. But you can compute the needed space yourself if you know your image size and frame rate. As ALFI only accepts 8 bit videos, the bit depth is given. We want uncompressed data. Assume you use a camera giving you 1920 px * 1200 px., at 25 frames per second. For one hour of video, you get 3600 s * 25 / s * 1920 * 1200 = 207 GB.
Q. Is there any sense to pixel binning to improve sensitivity?
A. Yes definitely, as this can increase sensitivity and the possible frame rate.
Q. What happens if cloud or haze comes along - do I keep on observing?
A. We suggest to keep recording, but note the time of the beginning of the interruption, and the end. In the log sheet, this is what the field "Record of events / Clouds in front of Moon" is for.
Q. How can lunar glare be minimised?
A. This is a topic in itself and would merit a dedicated workshop. It will include modifying your telescope - make sure to have dark, non-reflective interior tube coverings, large enough focusers to avoid internal reflections, and more. For now, we suggest you just use what you have. You might also try to move the illuminated part of the Moon outside the field of view, of course, keeping the Earthshine part in. Rotating the camera to avoid straylight from the secondary holder might also help.
Q. What proportion of flashes detected are lunar or of some other origin?
A. We cannot give a precise proportion, but expect lots of false detections from cosmic rays or sunlit peaks at the terminator. We can confirm a flash by having multiple observers recoding it at the same time.
Q. If I am first to report a flash do I get the credit of being the discoverer?
A. In our database we willlist the order in which observers report the same flash, but all observers deserve credit if they spot a flash. Other than for asteroid discoveries, there are no clear rules (yet!) for giving credit to impact flash observers. This campaign might help to come up with a solution.
Q. If I discover a flash or just send in confirmation reports, do I get a mention as an author or at least acknowledged in a paper?
A. We are planning to publish a paper about this campaign, depending on the results we will get. We still have to determine the policy for authorship or acknowledgement. We are planning an inclusive strategy - anybody who participated in the campaign (e.g. by registering and submitting observations, even if negative) will be included at least in the acknowledgments.
Q. Any advice on safety when observing in the early hours of the morning?
A. Wrap up warm. Warn the neighbours the day before. Be careful and aware of your surroundings. Watch out for cables.
Q. Should I be using a filter?
A. If you wish but as this is the first experimental observing session, try observing with white light as this maximises sensitivity.
Q. If the lower limit on the frame rate is 20 fps, what is the upper limit you would suggest?
A. The lower limit of 20 fps is a recommendation. LUMIO will use 14 fps. Thus, even if you cannot go to 20 fps, participate anyway. As for the highest frame rate:. Going higher means that we can better resolve the temporal distribution of the light emission. But with the resulting shorter exposure times, you will also reduce your sensitivity, so maybe don't go higher than 60 fps.
Q. How many minutes recording should I do?
A. Use approximately 5 or 10 min max to make the data analysis with the ALFI software easier.
Q. What happens to the frame rate if my bujer on the camera gets filled?
A. It will slow down your frame rate. Experimentation will show how many min you can record before this happens. If possible use a high rate USB.
Q. Can I use a colour camera if I don't own a monochrome one?
A. Yes but it is typically a factor of three less sensitive than a monochrome camera. However we could still use your data and it might constrain temperature of the flash. For the latter, however, a good calibration of the camera is needed.
Q. There seems to be a lot of emphasis on VirtualDub in the lecture about ALFI, why is this?
A. We have found that if problems are going to occur its because people haven't given the right prefix to the 8 Bit BMP files, or not used the right number of digits in the file name or did not state Monochrome (Y8). When AFLI was written the author wanted to use C for speed and although expert on programming algorithms, wasn't up-to-speed on reading AVI files direct from C.
Q. Can ALFI deal with non-tracking views of earthshine as not everybody has a tracking scope.
A. ALFI is better at this than the other two programs out there.
Q. How do I tell the dijerence between a cosmic ray and a LIF?
A. The former tends to be single/very few pixel events, whereas LIFs are more blurred due to the optics and seeing ejects. Make sure you record some stars with the same setup, to allow measuring the 'blurriness'.
Q. Will ALFI handle colour video or 12 bit video?
A. No - use VirtualDub to convert into 8 bit. Once a flash is detected we may ask for the original section of your video around the flash in colour (if available) or at 12 or 16 bits - if that is what you recorded it at
Q. What do the border controls do on ALFI?
A. They allow the user to add a margin around the border area in which ALFI will not look - we did this as old PAL/NTSC video has noise near the image edges. This functionality can also be used to avoid the detection of time stamps near the edge of the images.