Dear Exoplanets Team,
Just recently I posted my first transit.
The conditions were not ideal (average FWHM ~ 5), but I think it worked. The patrol is a bit short, 4 hours, but I think the data is good. A model seems close to the known one for this object.
On September 18, I'm likely to observe a transit of WASP-33. If any of you will also be following the same target, we can join forces. I am in Bulgaria (GMT+2), imaging with a Newtonian reflector 200/1000 mm, CMOS camera, and photometric V filter. I read that it is recommended to use a Red filter. I don't have one photometric, but I have a regular R filter for astrophotography and a narrow band Ha (12nm). Do you recommend the use of such filters? I have seen some people use a Ha filter.
Clear skies!
Very nice work, Nikola. Your results look very good. Sorry I am just now seeing your comment. How did the WASP-33 transit go?
If you use an R filter, it should be photometric because those are standardized.
Clear skies!
I'm…
Thank you, Anthony!
I'm happy to be part of your team now.
I managed to capture WASP-33. I did 5 hours of patrol, but the material now is difficult. The comp stars are too faint > 13 mag. I had to reduce the exposure to 8 seconds because the target star was too bright and easily saturated. Now I don't have a good signal on comp stars and de facto I only have one good comparison star of 9 mag.
Tough job, but I'll try to get data nonetheless. I hope to show a result soon.
Clear skies!
Here it is, WASP-33b!
With only one comparison star - 96. The others had too low SNR.
I notice that as the meridian approaches tot_C_cnts drops dramatically and accordingly the counters of the stars jump up. I think it is natural. What do you think? Is the problem with me? However, the telescope is quite amateurish. I plan to buy a completely new better telescope with the same 200mm aperture soon.
Regards!
Where on the plot was it crossing the meridian? I think WASP-33 crosses the meridian at about 3:30 am local time, so it's possible that the sunrise is affecting it.Also, the transit duration is a little under 3 hours for WASP-12 and it appears that this is a partial transit with of at least 4 hours.
The comps and target should give higher counts at the meridian.
The quality of your data points looks good. What reduction program are you using?
The meridian flip is after the transit, right at the end of data points, it is not included into the dataset. Most likely the scope makes problems when approaching the meridian, when its position becomes almost vertical and is close to the zenith.
What was the local time of the mid-transit time?
The local time of the mid-transit was 2:02.
I don't know why I cannot upload images here, but here are some pictures. The software I'm using is AIJ.
Transit Light Curve
Calculation from Model
Looking at individual images, I see that at times high thin clouds have passed over, changing the background barely noticeably.
I cannot agree more with what Gary Shaw says. Dennis's AIJ course can't be recommended enough.
But from the way you have presented your data, there does appear to be a correlation between the decrease of the tot_C_cnts and the increase of the rel flux of the target. That might imply that the comp star(s) may not be ideal.
I think I have problems with the calibration. I will reprocess the files again.
P.S. Not calibration, but clouds :(
Hi, Nikola. Looking at your updated plots, you appear to be on the right track. Try running AIJ again with just:
rel_flux_T1_n (AIRMASS detrended with transit fit) (normalized)
rel_flux_T1_n Transit Model
rel_flux_T1 (AIRMASS detrended) (normalized)
AIRMASS ppt (arbitrarily scaled and shifted)
That will give you some basics to get an overview of what is going on, then you can use other analyses to help you see what is really going on.
The…
Thank you, Anthony,
The root cause was the bad seeing. I'm glad because it doesn't mean I'm doing something wrong. :)
Last night was good and I had an opportunity to catch HAT-P-32 transit again.
A beautiful Light Curve came easily this time!
Nikola
The data appears pretty good and the format of the light curve is fine. I believe the preferred size and proportion of the page is 800 by 1000. This can be set on the Multiplot Main screen. The larger, more vertical format will give you room to better locate the data plots. Remember, it’s preferred to locate the top, raw data, plot referenced to the 1.0 line on the vertical axis.
Nice work!
Gary
If you haven’t…
Hi Nickola
If you haven’t already, you should carefully read the ‘Guide’ prepared by Dennis Conti, head of this Section. In it he shows how to present the Lightcurve and what parameters and information needs to be shown. likewise relative to the data and fit analysis page.
After you have digested the Guide, it’ll also help you with formatting the lightcurve and carrying out the analysis utilizing more of the features of AIJ. Several quick items to be aware of on the lightcurve is the requirement to include a plot of the raw data without detrending. As you’ll see in the Guide this data is presented on the top of that document. Also be sure to include all 6 of the customary ‘systematics’ at the bottom. See the example of these on the sample analysis (of WASP-12) depicted in the Guide.
Regarding the data fit panel, be sure not to include more than 2 detrend parameters - unless your observation involved a meridian flip, in which case always use that as one of just 3 detrends Are you aware of and using the ‘auto-optimization’ feature of AIJ? If not, be sure to learn to use it as it will be a huge benefit in automating the optimization process and lowering your RMS and BIC values- plus, it saves you tons of time.
If you’re interested in learning more about observing and analyzing transits, consider taking Dennis’ course - it’s a huge help in understanding the process and getting goods observations that are suitable for reporting. It’s great to see your observation data and I hope some of the above is helpful.
cheers, Gary
Thanks for the recommendations, Gary! I read the guide and will sign up for the course you recommend when the opportunity arises.
On re-analysis, it turned out that there had been a lot of thin clouds overnight, which introduced a lot of noise into the data and distorted the results.
After filtering out most of the faulty frames, the following pattern easily emerged:
Light Curve
Model
Here is the updated result.
Unfortunately, after cleaning the bad frames, I've lost the data reaching the egress and the transit duration doesn't fit the known model.
Today, Sep.22 I'm going to catch a transit of HAT-P-32 again!
Here are some suggestions:
Hi Nikola
As a follow-up to my note above, I've attached a copy of one of my lightcurve with 3 comments:
1. On the second line of the title, its preferred to indicate the aperture used to analyze the data in the lightcurve. This example shows the way to do that.
2. Once you have your initial lightcurve and have run the optimization with detrending to get the lowest possible RMS and BICon the Fitpanel, you should adjust the left and right 'Markers' to align with the fitted ingress and egress points on your transit fit curve. This will likely further decrease RMS and it will give you the 'observed' Tc of the fitted curve. From this you can determine how much the center of your 'observed' transit data diverged from the 'predicted' center of the transit time.
3. At the lower right of the Lightcurve page, the current version of AIJ is printed for record. Be sure to update AIJ before each use. The application has fairly regular updates so the latest version should always be used.
Hope this is helpful...
Gary
https://drive.google.com/open?id=1-YgIg6naKIIjTrZ4MBf33pbpJ5XxAi_p&usp=drive_fs
Thanks so much for the recommendations, Gary! I will take them into consideration.
Hello, Gary,
I just created a Google Photos album where I will put the Light curves from observations.
Also, I deleted WASP-33, because of the poor data quality.
I've updated the other HAT-P-32 transits so the old ones don't exist anymore in the AAVSO.
Thanks for the help, it was priceless!