To contact us:      | E-mail:


14 February 2021


Survived 2020, the year of catastrophic bush fires, floods, storms and the global COVID-19 pandemic. The virus is still with us but under control in NSW. Following planetary photography all through October, the C14 was reconfigured to HyperStar f/1.9 setup and regular NEO follow up resumed. There wasn’t much action however for two reasons.  It was a cool start to the summer due the El Nina this year resulting in mostly cloudy skies and rain. The garden loves it and there is no danger of bushfires, thank God and the El Nina. The other reason was the gradually reduction in the number of reachable targets candidates. The great majority of the NEO candidates on the NEOCP page are now fainter than 20V magnitude so unreachable to the C14.  The resulting MPECs  were 1 for November, 3 for December,  4 for January and 4 so far for February. 

29 August 2021


COVID-19 is still with us and raging out of control in NSW. We are severely locked down and restricted to moving in a 5km radius. Luckily the restrictions have no effect on the work of Arcadia Observatory.   New NEO asteroid follow up observations continued without interruption contributing to 45 NEO discoveries and their corresponding MPECs in the last 6 month.


In June, the famous Amor NEO 433 Eros was in a favourable position and I took the opportunity to observe it over two consecutive nights and generate two very good light curves with surprisingly large amplitudes of 0.72 V and a period of 5.27 +-0.01 hours, consistent with the accepted value.

6 March 2022


It has been a miserable summer in Arcadia.  Due to the still active La Nina there have been very few clear nights for observation. Both January and February only produced only one NEO confirmations. I did manage a remarkable observation however.


The James Webb Space Telescope (James Webb Space Telescope ( was launched on Christmas day last year to Le Grangian point L2.  I wondered shortly after the launch of JWST if it could be imaged at L2 1.5 million km away.  I was guessing that when fully deployed with its tennis court size sun shade, fully lit facing the Earth, the JWST would present a very bright target and should be detectable. Julian on the IceinSpace ‘general chat’ forum produced a plausible rough estimate of its apparent brightness of 14 to 15V at L2. It would always be in the night sky, somewhere on the meridian at midnight. This is not strictly true because JWST is not actually at L2 but orbits around it on a huge halo orbit that takes it more then 10 degrees away from it in the sky as seen from Earth. So a source of accurate ephemerides was needed and found at At the first clear night opportunity, which came a full month later on 9 February when the JWST was already at L2 and fully deployed. Pointing at the predicted position, I took 30 one minute images using my usual setup I use on survey run to find new moving objects. One minute exposure reaches to at least magnitude 18V on the C14 350mm scope and SBIG ST8300M camera so I was confident I would see it even without stacking. Running Astometrica’s ‘Moving Object Detection’ feature there were 2 moving objects in the field. One was recognised as asteroid 11478 the other, at almost he dead centre of the field was not recognised as a known object by the software so I’m certain it is the JWST shining at magnitude 17.6V. If this would have been a routine survey run I would have been getting excited but alas I expected exactly what was there. The image below shows detail of 5 stacked images generated by Astrometrica’s ‘track and stack’ function.


The day after I imaged JWST the skies closed again and lightning during an electrical thunderstorm ‘took out’ a large proportion of the electronic equipment in the observatory including the C14 hand controller, the telescope controller and the dome controller. Luckily I had spare parts for the faulty components but spent 3 days repairing everything. There is still some doubt about the health of the observatory computer as well as the integrity of the whole system. The rain has set in and there has been no opportunity for any actual observing.

The image on left shows detail of 5 stacked images generated by Astrometrica’s ‘track and stack’ function.









The image below shows two moving objects, the brighter one is asteroid 11478 the faint one in the centre is the JWST in orbit around L2


17 October 2023



This month I found Tycho Tracker thanks to Tamas Elek, fellow asteroid chaser at M38 in Hungary.  Tamas contacted me via this website and while we were discussing  astrometry software he pointed out that he uses Tycho Tracker by Daniel Parrot  ( for both NEO follow up astrometry and for searching for new objects. While I have been using Astrometrica ( for years very successfully for NEO follow up work I have had no luck with my modest but regular sky surveys. My problem with Astrometrica for survey work is that you can't use its brilliant 'track and stack'  function for searching for new moving objects simply because you have to know the 'motion vector' ie the direction an speed of your target. For NEO follow up the motion vector is published on the NOECP page along with its ephemeris. Using stacking you can reach objects as faint as 20V even with a modest telescope. For surveying you are limited to using single exposures which restricts you to looking for slow moving objects brighter than 18.5V using my C14.  Another problem I have with Astrometrica for survey work is that it requires very good quality images (perfect focus, good seeing) for its 'Moving Object Detection' to be useful, otherwise it gives you too many false detections and missed real movers. As I normally image ten 1.5 x 1.0 degree fields an hour, blinking these huge 3326 x 2504 pixel fields manually is extremely laborious and often makes me fall asleep as well as being unreliable.


The newish Tycho-Tracker features what is called Synthetic Tracking where the software tries a huge number of motion vectors in a range specified by the user while looking for stacked movers. To me it sounded like a brilliant idea but soon found the catch, it requires a computer with a powerful GPU (Graphic Processor Unit) otherwise it may take hours to process a single field. You can reduce the processing time greatly by restricting the range of motion vectors. At the limit, if you know the motion of an object, as in the case of objects on the NEOCP page you can use it as a NEO confirmation tool.  I'm still exploring its features but I already like the way it processes image data virtually automatically including image alignment and plate solving without having to enter accurate sky coordinates of your images.


I'm looking to change the algorithm in my automated survey software to take advantage of the features in Tycho Tracker.


So far 2023 has been 'peaceful' in terms of the weather comp;ared to that last two or three years when large electrical storms caused a lot of damage in the observatory.  It's El Nina time in Sydney and now I;'m worried about bush fires.


September was a record for successful NEO confirmations with the addition of 16 new MPECs to the score.

18 January 2024


Australian east coast weather is supposed to be under the influence of an El Nino this summer but it is looking more and more like a wet El Nina we had over the last 2 or three years. That is, there has been a lot of rain and very few clear sky opportunities for observing. It gave me a chance to progress two projects I have been planning.


To incorporate the recently discovered Tycho Tracker in my automated survey ‘pipeline’ I needed to restructure the software responsible for the imaging sequence so it will handle multiple consecutive images for each field when running a multi-field strip or rectangular area. Currently, when I use Astrometrica for processing, I run a ten field session where each field is imaged for 60 secs, three times in an hour, that is 20 minutes between imaging the same field. This implies moving the telescope after each image is taken, allowing time for downloading each image, re-pointing the ‘scope to the next field and allowing time for the auto-guider camera to find and lock onto a guide star. For a 10-field session the scope returns to the first field 3 times in the hour so any moving objects can be detected. This scheme doesn’t work for the Tycho Tracker because it needs at least 15 images of each field for its Synthetic Tracking algorithm to work. In principle what is required to achieve similar coverage in area and exposure time when using Tycho Tracker’s Synthetic Tracking feature is to divide up the total exposure time of 3x60 seconds between 15 exposures. That is 15x12 sec exposures for each field once, and not returning to the same field without three times. The problem with this calculation is that 15x12 second consecutive exposures is not long enough is, not long enough time for the Synthetic Tracking to detect slow moving objects. I have now modified the software so as to allow flexibility, via the session script, to chose either method and able to fine tune to optimise the session according to different criteria such as limiting magnitude and speed of moving objects


The second project I have just started is the design and implementation for a replacement of the electronics currently interfacing my vintage (CA 1990) Celestron C14 with the observatory computer. The current ‘Telescope Controller’ box is based on a microcomputer/microcontroller (Basic Stamp 2) that is no longer available as a direct replacement from the manufacturer so if it fails Arcadia Observatory is out of business. It has already failed once (result of a lightning strike close by) and I used up the only spare. I now want to redesign and build a replacement,  based on the very popular Arduino Mega microcomputer which I’m already using to control the observatory dome rotation. It’s early days but I have already started writing the code.