Fireball Tracking System Analysis

Analysis of the AMS Citizen Science Based Fireball Tracking System

Add your comments.

Written and Prepared By Mike Hankey March 18th, 2013
Contributors: Bob Lunsford – David Meisel, PhD – Vincent Perlerin, PhD

PDF Version –  Raw data used in this study in Excel Format

A Google Earth KML file containing the estimated trajectories of the significant AMS events from 2005-2013 can be download here: Trajectory Estimates for Significant AMS Events 2005-2013.  (KML file provided in zip format and must be unzipped. File can only be used with Google Earth. You can download Google Earth for free here.)


The American Meteor Society (AMS) was founded in 1911 and pioneered the study of visual meteor observations. The AMS has accepted reports of bright fireballs and bolides from the public since inception and online since 2005. All online reports are analyzed, approved, and grouped into events based on the location of the witnesses and the time of the sighting. The events are published on the AMS website along with a map displaying the witness location, sighting vectors and other details. Usually when enough reports are submitted, a rough trajectory of the fireball meteor can be plotted (see Appendix). Some of the fireballs reported to the AMS are associated with meteor showers while others are the bolide type, originating from the asteroid belt and sometimes causing sonic booms and leaving meteorites. A small percentage of reports are associated with the re-entry of space debris. Another small percentage of reports are non-meteor sightings. The AMS fireball report is a free service offered to the public and scientific community. The purpose of this paper is to present a summary of the fireball reports logged with the AMS from January 1, 2005 through February 27th, 2013.

Data Overview

As the graph below illustrates, the number of reports filed with the AMS has been increasing over the last few years. The level of traffic on the site, the popularity of the site and improvements to the software are all contributing to more reports being filed. More people are able to find the site and more easily complete the report application. These factors contribute to an increase in witness reports being filed with the AMS.


Not only are more reports being submitted, but also more unique events are being created. The following is a graph of unique fireball events registered by the AMS each year from January 2005 through February 27, 2013.


Fireball Events submitted to AMS per Year from 2005 to 2013

Circumstances affecting the number of reports and events collected by the AMS

There are circumstances affecting the data collected by the AMS that should be understood to prevent misinterpretation regarding the change of fireballs reported.

  1. From January 2005 through December 2010 the AMS fireball reports system was based on the same data collection form and process created by Bob Lunsford in 2005.
  2. In December 2010 the AMS fireball reporting system was upgraded to a database google maps system developed by Mike Hankey and company.
  3. The new software made it much easier for witnesses to file reports and resulted in more witnesses successfully reporting events.
  4. Internet traffic on the AMS site has increased since 2005. In current years, more people are connected with mobile devices, laptops and computers than ever before. This leads to more people reporting fireballs.
  5. At least one huge event with over 100+ reports in 2009 was space trash, and there was at least 1 space trash event for 2012. The graph of 100+ events has not been adjusted to account for space trash.
  6. A percentage of reports received are cloud contrails, planes, sun dogs or phenomenon other than fireball meteors. These reports are deleted when detected and generally tend to be events reported by only one witness.
  7. In November 2012, the AMS fireball reporting software was upgraded again and this upgrade resulted in an immediate increase of reports received. It is believed that improvements to the user interface lessened the amount of form abandonment and other training issues that may have kept witnesses from successfully logging events on the previous version.

Fireball Event Classes

The AMS has recently established a method of classifying fireball events to indicate the relative size of the event. This is a simple process based solely on the number of witnesses who report the same event. The logic for this approach is based on the belief that the more noticeable the event, the more people will see it and thus report it to the AMS. Of course this approach will not hold true for events that happen in under populated areas or odd times of night since fewer possible viewers would be present. The goal of the event classification is to add a level of differentiation between fireball events. Clearly an event reported by one person is less significant than an event reported by 100 people (regardless of how brilliant both fireballs may be). The frequency of these larger events will also be significantly less, i.e. one can agree that 100+ witness events will be less common than 1-witness events.

As total traffic to the site increases, the total number of witness reports for any given event will also increase, so viewing the data for various sized events can also be misleading. However, viewing the events by class over time provides context to the data collected by the AMS and the events the society has logged over the last 8 years.

The AMS has adopted a classification system for each event in an attempt to quantify the size of the event. The fireball classification system the AMS has chosen to adopt for this purpose follows:

Event Class Number of Witnesses  Description
Unconfirmed Event reported by only 1 witness. The vast majority (70-80%) of events reported to the AMS are classified as unconfirmed.
Confirmed Events reported with 2-5 witnesses are classified as confirmed events.
Mid Sized Events with 6-20 witness reports are classed as mid-sized events.
Significant Events with 21-50 witness reports are classed as significant.
Large Events with 50-100 witness reports as large events.
Huge Events with 100+ witness reports are rare and classed as huge events.
Sonic A special class for events with sonic boom effects has been created to classify deep penetrating fireballs. 2 or more witnesses reporting a delayed boom are required to classify an event as sonic.

Below is a data table representing all of the AMS reports by event class from January 2005 through Feb 27, 2013.

Events By Class Per Year
1 Witness Report (unconfirmed)
2-5 Witness Reports (confirmed)
6-20 Witness Reports (mid-sized)
21-50 Witness Reports (significant)
51-100 Witness Reports (large)
100 or more witness Reports (huge)
Events with Confirmed Sounds (sonic)
Total Events Per Year

Each row in the table above shows the total fireball events recorded with the AMS by classification and year.

Graphs of AMS Fireball Events

Graphs for this data table have been split to compensate for scale. Below are graphs for the first 3 classes of fireball events.

Fireball Events for class Unconfirmed, Confirmed and Mid-Sized submitted to AMS per Year from 2005 to 2012 (with Jan-Feb 2013)

The next graph shows the larger event classes by year.

Fireball Events for class Significant, Large and Huge submitted to AMS
per Year from 2005 to 2012 (with Jan-Feb 2013)

Since the program’s inception in 2005,the AMS has also logged flags to indicate delayed sonic booms
associated with fireball events. Sonic booms sometimes accompany large bolide events when the fireball
detonates and fragments in a terminal burst, often leaving meteorites. Events with sonic booms are
considered rare and a small percentage of reports with sonic booms are submitted each year. AMS events
with sonic booms have risen since 2009 and doubled in 2012.

The graph below charts the sonic boom events reported by 2 and 3 witnesses.

Fireball Events with at least to sonic reports per Year from 2005 to 2012 (with Jan-Feb 2013)

Certainly fireball events with large numbers of witnesses are most noteworthy. The AMS has only classified 14 events as “huge” (100+ witnesses) in the 8 years of operations. 6 of those events were logged in 2012.

Huge Fireball per Year from 2005 to 2012 (with Jan-Feb 2013)

When we look at a graph of huge events by month, we can see one huge event was reported as far back as 2005, but also a clustering of huge events was reported in 2012. We can also note that at least 4 huge events were reported to the AMS before the software upgrade at the end of 2010 (indicating that the early version of the software had enough reach and capability to register huge events).

Huge Fireball per Month from Jan 2005 to Feb 2013

As we can see from the graph above, the frequency of 100+ events reported to the AMS increased leading up to and including 2012.

Below is a graph of meteorite recoveries from witnessed falls from 2000 – 2012.

Meteorite Recoveries from Witnessed Falls
per Year from 2002 to 2012 – USA vs Rest of the Word

While not grossly out of proportion with past years, of coincidental interest is the fact that more meteorites from witnessed falls were recovered in 2012 than in the past 12 years. The running 12-year average for 2000-2012 is 7 meteorite recoveries per year worldwide. In 2012 there were 12.

Recommendations Regarding the Frequency of Fireballs Reported to the AMS

The AMS fireball reporting tool, while a useful service to the public and scientific community, is not a controlled scientific environment that one can use to reach conclusions about the frequency of fireballs. The AMS in no way suggests that the frequency of fireballs has increased or is increasing. This is not to say that fireball and bolide rates are not increasing, they may in fact be increasing, we simply cannot prove this with the AMS reports alone. We can only conclude from the data collected that more users are submitting fireball reports. It is understood that the rates of fireballs will increase and decrease year to year, but in order to answer the question “Have the rates of fireballs been increasing lately?” a more comprehensive study is needed.

Analysis of DOD & DOE Bolide Data

As documented in, The Flux of small near-Earth objects colliding with Earth (Letters to Nature – vol420, 2002), the United States Department of Defense and Department of Energy operate space-based systems that are capable of detecting bolide events across the entire globe. From the period of February 1994 to September 2002 bolide data from this system was analyzed, and based on that analysis these conclusions were made:

We estimate that the Earth is on average struck annually by an object of energy, 5 kton (with a possible range of 2–10 kton), and struck each month by an object with 0.3 kton of energy. Every ten years, an object of energy,50 kton strikes Earth.

Letters to Nature – The Flux of small near-Earth objects colliding with Earth, P. Brown, R.E. Spalding, D.O. ReVelle, E. Tagllaferri & S.P. Worden NATURE | VOL 420 | 21 NOVEMBER 2002

Re-executing this 2002 study using the DOD bolide reports from 2003-2012 would provide definitive insight into the recent perceived increase in fireball rates. Another benefit resulting from this study would be the development of a worldwide large-bolide frequency table covering a 20-year time period. It appears the DOD agrees more study of fireballs is needed and they have recently stated they will re-enable the sharing of bolide data with the public, as reported by Leonard David at on February 26th, 2013.

On March 14th NASA announced the release of a Fireball and Bolide website where the analyzed results of the DOD data will be published. The AMS hopes NASA will give attention to the most significant bolide events that were recorded over the USA and reported to the AMS over the last five years. The release of bolide data relating to these events will help the AMS calibrate the reporting of our system so that future analysis will be more accurate. We will also be able to use this data to better understand past events and reach more meaningful conclusions about the size, locations and origins of these asteroids and meteoroids that have already struck Earth.


Since 2005, the AMS fireball report system has logged over 17,000 reports, identified more than 8,000 unique fireball events and been successfully used for different purposes by the scientific community and various government agencies including NASA, the Coast Guard, and the Air Force. Several meteorite recoveries have occurred in part due to the data collected by the AMS and the AMS fireball reports have assisted meteor and meteorite research at NASA Ames and other NASA offices. The re-entry of satellites and space debris has been confirmed by AMS reports. The Coast Guard has also used the AMS reports to vet calls about crashed airplanes off the coast of Florida (which later turn out to be fireball meteors). The AMS continues to improve the systems that collect and analyze the fireball report data. We have developed trajectory analysis programs that automatically determine the flight path of fireball meteors (within a margin of error) and we are in the process of creating orbital analysis programs for the improvement of meteorite recovery and meteor research. The AMS fireball log is the most comprehensive set of public data regarding bolide and fireball events that have taken place over the United States since 2005.

While the AMS fireball log has many uses and benefits, based on the AMS reports alone it is not possible to make conclusions about an increase or decrease in fireball/bolide events from year to year. However, the data shows that reports submitted to the society have been increasing and a significant increase in large events was specifically noticed in 2012. This warrants further study. A pairing of DOD bolide data with AMS event data would prove useful to the scientific research of meteors striking Earth. In addition, the AMS reports identify the date, time, location and relative size of significant fireball events. This information could be useful in pinpointing events inside the DOD’s data sets. The AMS makes all report data available through our website and encourages the use of our data.

The AMS appreciates criticism and review of this analysis and the fireball data presented herein. Anyone with questions regarding this article or a desire to analyze the AMS raw data for statistical purposes may contact Mike Hankey regarding their requests. Mike can be reached at