July 20, 2015
For the first few weeks of the summer, I used an antenna hooked up to a raspberry pi located in Tucson to receive transmissions from local airplanes. I used these data to write a python script to transform the Lat/Lon and elevation info of the planes into Alt/Az and then x/y/z (for the all-sky camera).
Mike Warner sent a zip of data he collected with the antenna already down there, and I ran it through my code for a sanity check.
Each plane has its own Hex ID so the different colors represent the different planes flying by. I'm going to work on setting up a python notebook because the 3D images are so much more fun to play with in interactive mode.
Today Chuck and I got dump1090 running on the Mac Mini in Cerra Pachon and started collecting data so we can look at the airplane frequency of appearance.
July 28, 2015 - Plots!
I've been cranking away at making some plots to look at how busy the sky is.
All of the following plots were generated using data from the past 24 hours.
Number of signals received above 10, 30, 45, and 60 degrees.
2D histogram of all the data using the x, y-coordinates that match the All Sky camera.
The units x, y units are a little strange. The plane positions are plotted on a circle with a radius of 1. It's what you would see if you looked up at the sky from the antenna (so what the All Sky Camera is seeing). The color represents how many signals are received over a certain area (the binning is 60 by 60 so the detail of the flights shows up better). There are some definite patterns from the various flights that cross overhead. Patterns in flight is something I'm going to look into next.
2D histogram split into day and night hours. Day hours = 6 to18. Night hours = 18 to 6
I have yet to figure out how to make the scaling on the two plots the same (and if anyone has any tips they would be more than welcome!). The axes and binning are the same as the total raw data histogram.
The data from the previous plots are going to be used to compare aircraft and their contrails to the All Sky camera images, but the next few are in Altitude and Azimuth coordinates to view it a little differently.
Raw data in Alt/Az coordinates. Each color represents one plane.
From the Alt/Az plot (or even the Lat/Lon plot) you can see there are paths that the same plane follows multiple times within 24 hours. These paths are going to be the most likely places in the sky where contrails form.
2D histogram of all the data, Alt/Az
Here the binning is 73 by 19, so each box represents a 5x5 degree section of the sky.
August 7th, 2015 - REU Results
Below are the slides from my final presentation on my summer project. Note that the 4th slide is actually supposed to be a video but I wanted to save space with a pdf.
Chuck and I created a git repository you can look at here that contains all my code and data from this summer. I also have my final paper in there, which I'm still working on.
I also wanted to share how I collected the ads-b data with the Mac Mini.
cd to /Users/christopherstubbs/rose and enter the following command:
$ nohup ./dump1090/dump1090 --interactive --net # or whichever dump1090 you want to use. --net is necessary to access the TCP port
And, there might be a better way to do this, but I had to close the shell in order for dump1090 to run uninterrupted.
In the same directory enter:
$ nohup collecth.sh
And again close the shell. Now both dump1090 and the collect script will run uninterrupted and add to the out_collect.csv file.
October 10th 2015 - Contrail background/research plan
Last week Chris and I came up with a action plan for this semester. The goals for the end of this calendar year are:
- Research contrails and be able to predict their formation and persistence patterns above the LSST site
- Analyze images from the allsky camera and look at contrail evolution
I'm working on getting access to the NOAO computer system again so I can start collecting images, but in the mean time I'm researching contrail formation and persistence.
As of right now I've found two groups with papers/data relevant to the topic: Measurement of Ozone and Water Vapour on Airbus in-service Aircraft (MOZAIC) and the Subsonic Aircraft: Contrail and Cloud Effects Special Study (SUCCESS). Most of the motivation behind contrail research is from environmental or military research groups. The environmental research is dedicated to determining if anthropogenic cloud formation is significantly affecting the environment, whereas the military is researching to determine air conditions for which stealth aircraft can remain undetected.
I have yet to find a nice easy equation I can plug air conditions into and get probability of contrail formation as an output, but I have found the following:
- Temperature, relative humidity, and total atmospheric pressure are the environmental conditions that determine whether contrail formation is possible [Jenson, 1998]
- Temperatures for contrail formation are 0-2K below the liquid-saturation threshold temperature, and if ambient air is not supersaturated with respect to ice, contrail formation should not be possible above -45 to -50°C at typical pressure ranges of commercial aircraft cruise altitudes[Jenson, 1998].
- Formation has been recorded in the presence of ambient cirrus clouds. [Jenson, 1998].
- Most contrails persisted for no more than a few minutes. Dissipate when aircraft vortices became unstable [Jenson, 1998].
- At high temperatures and supersaturation, warm contrails form in the presence of cirrus clouds and persist longer[Jenson, 1998].
- Contrails can be distinguished from natural cirrus by their tendency to remain thin and generate strong laser backscattering and depolarization, even an hour or more after formation [Sassen, 1998]
- Contrails tend to be thin and strongly scattering cloud layers, which can generate coronas indicative of small particles even after persisting an hour or more [Sassen, 1998]
- Persistence requires only supersaturation relative to ice[Burkhardt, 2008]
- Contrails form typically below 233K and 300hPa[Burkhardt, 2008].
I'll keep looking for papers that can give more precise temperatures, pressures, and humidities that contrails form and persist at.
December 5th, 2015 - Contrail Surface Brightness Profile
On 11/14 and 11/15 Chris took images using the All Sky Camera to capture contrail evolution across the sky. Today we analyzed the contrail traveling through images 400-480 from 11/15. Chris wrote a beautiful MATLAB code that acts like IRAF's imexam, except better because it allows you to look at the brightness across a diagonal (excellent for contrails curved relative to the camera).
We analyzed the one contrail and I know have a nice set of data to check out.
Example of a surface brightness profile from Chris' code. Note the double peak from the two different contrail streams.
I'm going to use these date look at the following features of the contrail with respect to time:
- movement across the chip
- attenuation
- change in thickness
Hopefully I'll be able to get a nice plot I can put on my AAS poster!
We also took a few hundred images today while the dump1090 program was running to have complementary data, however there were no contrails today, which prompts me to do a little more background reading on their formation.
One of the next steps is for me to write a code that starts taking images when aircraft fly within a certain distance of the site, so we can start collecting data for further analysis.