Welcome to the Global Meteor Network's wiki page!
The Global Meteor Network (GMN) is a world wide organization of amateur and professional astronomers alike, whose goal is to observe the night sky using low-light video cameras and produce meteor trajectories in a coordinated manner. Here you will find information on the purpose and structure of the GMN, assembling and operating your own meteor camera, contributing to the development of RMS the GMN software, as well as information on how your observations as a citizen scientist can contribute to the further understanding of our solar system's formation and evolution.
If you have come here to find out how to build and setup a camera from scratch, jump ahead to this section !
Global Meteor Network Overview
Meteor Detection Station
What is an RMS GMN station?
- A RMS-based GMN station that is the subject of this Wiki consists of a Raspberry Pi (RPi) single board computer, a low light level security video camera, and the RMS software package. The camera is securely mounted in a weatherproof housing, pointed at the sky, and connected to the RPi with a POE (Power Over Ethernet) cable. The RPi is connected to the Internet via WiFi, and to be a part of GMN network, you’ll need a fairly powerful Raspberry Pi (RPi 3B+, RPi 4 or better) and a reasonably fast Internet connection. The internet connection is primarily required to enable data upload to a central server each morning as well as provide automatic updates for the RMS software.
- Nightly, the RPi starts recording video from the camera shortly after local sunset continuously compressing and storing the video data locally. Each morning before sunrise, after capture is complete, the RPi analyzes the video and extracts your nightly station’s meteor observations. These extracted video “clips” of detected meteors are then archived and uploaded to a server. The clips can total hundreds of megabytes on a “busy” night (e.g., in a heavy meteor shower, or a night with a lot of false detections--progress is being made on the detection software). The server finds meteors which were observed with more than one station and this enables the server to triangulate the meteor trails in 3D and calculate the orbits of the meteors.
What do I need?
You'll need a Raspberry Pi with the software on, and a camera kit. We strongly recommend the Pi4 model. The software will run on a Pi3 but it is much slower and now also not maintained option. The Shopping list with everything you will need can be found on this page.
It is also possible to run multiple cameras on a Linux PC. More details here.
How do I obtain a camera?
There are two options:
Buy a Camera
You can buy a camera and Pi prebuilt and ready to install. These are available from a couple of suppliers. The Croatian Meteor Network sell prebuilt cameras as explained on this page. Alternatively, if you're in the UK, you can obtain cameras from the UK Meteor Network
Build your own from scratch
This requires some basic DIY skills and some familiarity with the Raspberry PI, but don't be put off. The instructions are comprehensive and if you get stuck, you can ask for advice in the groups.io forum.
Click on this link if you want to build a camera from scratch.
Advanced RMS installations and Multi-camera support
If you would like to explore some advanced possibilities to install RMS software onto various platforms or to run multiple cameras on a single Linux computer, please have a look at this page. In case you are planning to run RMS software on the Raspberry Pi 4, then please use the supported and easiest possibility by using the image we have prepared for you. More information is in this extensive guide.
Can I use a commercial all-sky camera?
- Generally no due to the lack of sensitivity. But see this recent experiment
Operating and maintaining your GMN station
- Please note that GMS is a nascent operation and you may share some growing pains if you choose to be involved -- we're still working out some bugs and making improvements here, which may be an opportunity to help if you have programming skills! ;-) So note that the workload of day-to-day operation can be non-zero, and might take a little bit of your time.
- Ideally, you'll want to monitor your RMS RPi system(s) daily to look for freezes or glitches or other problems... like birds nesting or soiling the camera window, people accidentally unplugging the power cord, mice (or cats or dogs!) chewing on the camera Ethernet cable, etc.
- Although we are getting close, this is not a "power up and forget about it" system yet.
- However, by its very nature, the GMS network is inhabited by a lot of people who are willing to help newcomers getting started. So, here are some clues for daily operation of your RMS camera.
So what does the meteor camera do over the course of 24 hours?
- The RMS Python based system calculates the sundown to sunrise interval and schedules video camera capture all night long. Depending on the video camera and capabilities of the RPi, the camera captures 25 or more frames per second between evening and morning twilight. During the continuous image capture, the station begins processing captured image data, doing a pre-screening to target frames with a suitable number of stars (usually around 20) that makes it worth looking for meteor detections. Once data capture has finished, the station switches into processing all the promising frames for detections, then refining the astrometric accuracy of every positive detection. Using the station platepar (plate parameters) calibration file, processing iterates to find the best astrometry and photometry solution for each detected meteor. Once this process has analyzed each detection, summary files are created.
- These summary files include text file data presentation in several widely accepted formats (CAMS and UFOorbit), as well as graphic plots of detection frequencies throughout the night, a set of thumbnail images of detections, a set of thumbnail images of data captured throughout the night, a single image with all detections stacked together, plots of photometry, astrometry, and camera pointing drift in arc minutes throughout the course of the night as the mount or building flexes, a flat file for correcting images, and a plot of all detections showing any identified radiants. Finally all results are combined into a single compressed archive, which is automatically uploaded each morning to the central server. Optionally, you can create a mp4 movie showing a time lapse of the night’s captured images.
- Each morning you can review the result files on the RPi, and copy anything you want to your computer or tablet.
- Your primary scientific data is automatically uploaded to the central server every morning when data processing is done. However once it has done this, RMS will purge out the oldest data to free up space for the next night's run.
- So, you may want to copy some of the data to a PC, NAS or cloud for further analysis of your own. The data you should consider backing up are the contents of ~/RMS_data/ArchivedFiles, which holds the individual files and data that RMS determined were probably meteors. Full detail on how to nbare beyond the scope of the GMN Wiki, but tools such as robocopy (for Windows) and rsync(for Linux/MacOS) are ideal. These tools can 'mirror' data across a network. If you want help configuring these, ask in the Globalmeteornetwork group on groups.io.
- We've also built some automated tools that can help to back up any additional data to a thumb drive inserted into the RPi. Please ask in the group about these.
Backup and restore configuration and RSA keys
- Open a terminal and execute the command Scripts/RMS_Backup.sh. A compressed .zip file containing all important configuration files and keys will be created in user's home directory with the prefix RMS_Backup and .zip extension. Example: /home/pi/RMS_Backup_XX0001_2023-01-28.zip.
- Copy the .zip file to a safe place outside RPi, it will be useful later to restore the system in case of failure. Note it contains the RSA public and private keys used to contact GMN servers, keep it secret.
- To restore the configuration, unzip the backup file in some folder on the RPi and copy the files .config, platepar_cmn2010.cal and mask.bmp to the folder /home/pi/source/RMS/, and the files id_rsa and id_rsa.pub to the folder /home/pi/.ssh/ as in the following example:
- cp .config platepar_cmn2010.cal mask.bmp /home/pi/source/RMS/
- cp id_rsa id_rsa.pub /home/pi/.ssh/
- Make sure that RSA key files permission bits are correct by using the command:
- chmod 400 ~/.ssh/id_rsa*
Viewing the data
- To view the data, you can use CMN_binViewer software which is already installed in the RMS SD image.
- There is also a Windows version you can install.
- Important note : You can also open the images in astronomical FITS viewers such as FITS Liberator or Pixinsight, though the results may be surprising. For example in FITS Liberator, the image will be upside down. This is an artefact of how the software reads the image. In space, there's no 'up' or 'down' and so the FITS specification does not dictate whether the pixel (0,0) is at the bottom left or top left, or indeed one of the other corners. Some software, notably FITS Liberator, treats the top left as the origin and so terrestrial images will be displayed mirrored vertically.
Tools and Utilities
- RealVNC or AnyDesk remote connect tool allows station access from anywhere. Access from outside your network is enabled by use of an OpenVPN connection address available to meteor stations. Alternatively, with VNC and Teamviewer, you can create an account and team on their websites, and then remotely access your station.
- Samba data directory access, allows you to copy data results directly from your RPi to your computer or tablet.
- CMN_binViewer can be used to view standard fits image files containing meteor detections. It runs on the RPi, and is also available under Windows.
- UFO Orbit allows you to process data from multiple stations and generate unified radiants of two or more stations seeing the same meteor. It can plot the shared object ground path, orbital characteristics, and can output a summary file of all objects seen by more than one station, which can be used for further analysis.
- RMS software can be installed under Windows to allow much of the RMS python-based code to be executed on your computer, so it can be run against meteor station data you have transferred to your computer from the RPi.
- You can run RMS Python jobs on the RPi to sample the image files captured all night long and condense them into a mp4 movie. This creates a sometimes mesmerizing summary that can run for over 2 minutes in length for winter time data.
What can I do with my GMN station?
Using SkyFit2 to perform astrometric and photometric calibration + Manually reducing observations of fireballs and computing their trajectories:
Data analysis with SkyFit2
SkyFit2 is a program within the RMS library which supports analyzing optical meteor data in most optical formats that are in use today, including videos in any popular video format (mp4, avi, mkv), a sequence of static images, or a single image with shutter breaks.
This video tutorial explains how to using SkyFit2 to perform astrometric and photometric calibration on GMN data and manually reduce observations of fireballs, including computing their trajectories.
A more general and detailed description of SkyFit2 is given at the SkyFit2 page.
What should I back up when re-flashing an SD card or a USB disk?
- The .config, platepar and mask files that are in the RMS source directory, plus the whole contents of the hidden directory /home/pi/.ssh.
If your SD card or a USB disk fails or becomes corrupted, the config files can be fetched from the server as they are uploaded every day together with the data. However the contents of .ssh are essential for connection to the server, so you must also save these. Once you set up a new SD card or a USB disk, return the files in their original location.
What are the values in FTPdetectinfo_* file designated as hnr mle bin Pix/fm Rho Phi?
- Some of these values are not used in RMS (hnr mle bin), but they are in CAMS, so they are here to conform to the standard. Thus they are all zeros. The others are:
- - Pix/fm - Average angular speed of the meteor in pixels per frame.
- - Rho, Phi - Parameters that define the line of the meteor in polar coordinates, see here for more details. Rho is the distance of the line from the centre of the image, and phi is the angle of the line as measured from the positive direction of the Y axis (basically a line going from the center of the image to the top of the image), the positive angles are measured clockwise (I think, the CAMS standard might define these parameters a bit differently, the Y axis is flipped).
- The intensity is the sum of all pixel intensities of the meteor on a given frame. Let's say I represent an area around the meteor on a given frame like this, where the numbers are pixel intensities on an 8-bit image (so they can range from 0 to 255):
- and the pixels values inside the red boundary represent the meteor blob on the frame, the intensity would be the sum of all numbers inside the red boundary.
- This value is later used to compute the magnitude. The magnitude is computed as: mag = -2.5*log10(intensity sum) + photometric_offset. The photometric offset is estimated in SkyFit by fitting the line with slope -2.5 through pairs of known magnitudes of stars and logartihms of their pixel intensity sum. The photometric offset is basically the intercept of that line. The constant slope of -2.5 comes from the definition of stellar magnitudes.
GMN data policy
The Global Meteor Network produces several levles of data products:
- Level 1 - The lowest level data (i.e. as close to "raw" as possible) are the FF image and FR video files saved to the RPi by the capture code and the fireball detector.
- Level 2 - The meteor detector uses these data to extract positional and brightness information of individual meteors (FTPdetectinfo file), and images are also used for astrometric and photometric calibration (platepar file). Meteor and star detections are used to generate a range of plots suchs as the single-station shower association graph, camera drift graph, etc. The calibrated meteors measurements get uploaded to the GMN server together with the raw images of individual meteors.
- Level 3 - The software on the server correlates individual observations and computes multi-station meteor trajectories which are published daily on the GMN data website. This data is made public under the CC BY 4.0 license.
Operators of individual GMN stations exclusivery own the Level 1 and Level 2 data their stations produce. In practice, this means that they are free to share this data with other meteor networks if they wish to do so. The data that gets uploaded to the GMN server will not be shared publicly nor with other parties without the operator's consent, but may be used internally by the GMN coordinators to manually produce other data products (e.g. trajectory of a meteorite dropping fireball, analysis of a meteor shower). All station operators will be credited for their data in all GMN publications.
- The IstraStream.com is an independent hosting site which primarily hosts data by cameras sold by IstraStream, but they also host several other stations. As for mid-2023 Istrastream stopped taking new cameras and is aimed to be completely replaced by the GMN weblog.
This document explains what every plot on the IstraStream weblog means:
For More Information
This wiki has numerous FAQs and tips on maintaining, monitoring and managing your system, and several explainers such as how to calibrate and create a mask, how to copy data and so forth.
- Vida, D., Blaauw Erskine, R. C., Brown, P. G., Kambulow, J., Campbell-Brown, M., & Mazur, M. J. (2022). Computing optical meteor flux using global meteor network data. Monthly Notices of the Royal Astronomical Society, 515(2), 2322-2339.
- Moorhead, A. V., Clements, T., & Vida, D. (2021). Meteor shower radiant dispersions in Global Meteor Network data. Monthly Notices of the Royal Astronomical Society, 508(1), 326-339.
- Vida, D., Šegon, D., Gural, P. S., Brown, P. G., McIntyre, M. J., Dijkema, T. J., Pavletić, L., Kukić, P., Mazur, M.J., Eschman, P., Roggemans, P., Merlak, A., & Zubović, D. (2021). The Global Meteor Network–Methodology and first results. Monthly Notices of the Royal Astronomical Society, 506(4), 5046-5074.
- Moorhead, A. V., Clements, T. D., & Vida, D. (2020). Realistic gravitational focusing of meteoroid streams. Monthly Notices of the Royal Astronomical Society, 494(2), 2982-2994.
- Kukić, P., Gural, P., Vida, D., Šegon, D. & Merlak, A. (2018) Correction for meteor centroids observed using rolling shutter cameras. WGN, Journal of the International Meteor Organization, 46:5, 154-118.
- Vida, D., Mazur, M. J., Šegon, D., Kukić, P., & Merlak, A. (2018). Compressive strength of a skirting Daytime Arietid-first science results from low-cost Raspberry Pi-based meteor stations. WGN, Journal of the International Meteor Organization, 46, 113-118.
- Vida, D., Gural, P., Brown, P., Campbell-Brown, M., Wiegert, P. (2019) Estimating trajectories of meteors: an observational Monte Carlo approach - I. Theory. arXiv:1911.02979v4 [astro-ph.EP 21 Apr 2020]
- Vida, D., Gural, P., Brown, P., Campbell-Brown, M., Wiegert, P. (2019) Estimating trajectories of meteors: an observational Monte Carlo approach - II. Results. arXiv:1911.11734v1 [astro-ph.EP 26 Novr 2019]
- Vida, D., Mazur, M. J., Šegon, D., Zubović, D., Kukić, P., Parag, F., & Macan, A. (2018). First results of a Raspberry Pi based meteor camera system. WGN, Journal of the International Meteor Organization, 46, 71-78.
- Vida, D., Zubović, D., Šegon, D., Gural, P., & Cupec, R. (2016). Open-source meteor detection software for low-cost single-board computers. In Proceedings of the International Meteor Conference (IMC2016), Egmond, The Netherlands (pp. 2-5).