1 Introduction

Video observations of the Moon during the Leonid storm in 1999 (Dunham et al. 2000; Ortiz et al. 2000) confirmed that lunar meteoroid impacts are observable from the Earth. One probable Geminid impact was observed from lunar orbit by Apollo 17 astronaut Dr. Harrison Schmitt (NASA 1972). Since NASA’s Constellation Program, which will place crews on the lunar surface for up to 6 months at a time, is currently in the preliminary design stages a new lunar impact ejecta environment model is needed. This exposure time is vastly increased over the Apollo Program and the risk from meteoroid impact ejecta must be better understood so that shielding on lunar spacecraft, spacesuits, and surface systems can be optimally designed. The existing model, NASA SP-8013 (NASA 1969), shows ejecta at a given particle size to be 10,000 times as abundant as primary meteoroids. This violates conservation of energy and is probably overly conservative which will result in lunar spacecraft designs with too much meteoroid shielding and hence too much weight.

Since our organization, NASA Marshall Space Flight Center’s Natural Environments Branch, houses the Meteoroid Environment Office and the Constellation Program Environments and Constraints co-lead, we have the responsibility for defining the ejecta environment and have undertaken a program of observations, testing, and modeling to do so. Our first results were reported by Cooke et al. (2006, 2007).

2 Observational Technique

The observations are carried out at the Automated Lunar and Meteor Observatory located on-site at the Marshall Space Flight Center (latitude 34.66 N, longitude 86.66 W). The instruments consist of two Meade RCX400 14 inch (355 mm) diameter telescopes with Optec 0.33× focal reducers and StellaCam EX monochrome video cameras. The effective focal length is approximately 923 mm giving a horizontal field of view of 20 arc minutes covering approximately 4.5 × 106 km2 or 12% of the lunar surface (see Fig. 1). The limiting stellar magnitude at the 1/30 s frame rate is approximately 12. The video from the StellaCam EX is digitized using a Sony GV-D800 digital tape deck and sent by Firewire to a personal computer where it is recorded on the hard drive for subsequent analysis.

Fig. 1
figure 1

Camera field of view and orientation

Full size image

The observations are made of the earthshine portion of the moon when the sunlit portion is between 10% and 50% illuminated. This occurs on five nights and five mornings per month. We do not observe during phases less than 10% since the time between twilight and moon rise or set is too short. We do not observe during phases greater than 50% because the scattered light from the sunlit portion of the moon is too great and masks the fainter flashes. Large lunar features are easily visible in the earthshine and are used to determine the location of the impacts on the lunar surface.

The recorded video is analyzed using two custom programs. LunarScan (available at http://www.gvarros.com) was developed by Gural (2007) and modified to read the video files. The threshold for pixel exceedance is set to 3.5 times the standard deviation over the mean image. The mean and standard deviation are tracked on a frame by frame basis using a first order response filter for each pixel channel independently. The threshold exceedances are then examined using a spatial correlation filter that looks for a row containing an adjacent triplet of exceedances bordered two rows above or below by a pair of exceedances. The software finds flashes in the video which meet these criteria and presents them to a user who determines if they are cosmic ray impacts in the detector, sun glints from satellites between the Earth and the Moon, or actual meteoroid impacts. By requiring that a flash be simultaneously detected in both telescopes, cosmic rays and electronic noise can be ruled out. Some of the detected impacts were observed with only one telescope early in the program but only flashes which spanned more than two video frames and showed a proper light curve (abrupt brightness increase followed by gradual decay) were counted. There have also been two impacts independently observed by amateur astronomers using 8 inch (200 mm) telescopes (G. Varros, D. Clark private communication). For short flashes where satellite motion might not have been detectable, custom software was used to check for conjunctions with Earth orbiting satellites whose orbital elements are available in the unclassified satellite catalog (http://www.space-track.org). Since there is some probability that orbital debris or a classified satellite not listed in this catalog could cause such a short flash, another observing station has been constructed in northern Georgia about 100 km from MSFC. This will allow parallax discrimination between impact flashes and sun glints from manmade objects, even at geosynchronous altitude.

After detection and confirmation, another computer program, LunaCon, is used to perform photometric analysis (Swift et al. 2007). Background stars are used as photometric standards to determine the observed luminous energy of the flashes. Modifications to LunaCon to improve photometric calibration, determine observed lunar surface area (collecting area), and detection threshold are described in Swift et al. (2007).

3 Observational Results

A total of 54 impact flashes were observed between November 2005 and May 2007 (Fig. 2). These were observed in a total of approximately 190 h of observation. We assumed that impacts detected during the 3 days around the peaks of major meteor showers which were located on the portion of the Moon visible from the shower radiant (determined using LunarScan) were due to shower meteoroids. It is possible that sporadics caused some of these impacts but the rates increased so dramatically for the showers that it is likely they were actually shower meteoroids. Since the velocities and impact angles of shower meteoroids are well determined, we are currently following the technique of Rubio et al. (2000) to determine luminous efficiency using our Lyrid and Geminid impacts (11 impacts each). There were approximately 16 h of observation time during these shower periods and 27 possible shower impacts were seen giving approximately 1.6 flashes per hour.

Fig. 2
figure 2

Impact flashes observed between November 2005 and May 2007. Continuous monitoring was from April 2006 to May 2007. The yellow numbers are probably sporadics, the white is likely a Taurid, blue are Leonids, green are Geminids, and red are Lyrids. A complete list of candidate impacts is given at http://www.nasa.gov/centers/marshall/news/lunar/index.html

Full size image

There were coincidentally a total of 27 likely sporadics; 21 were observed on the western hemisphere of the Moon (waxing phase) and six on the eastern (waning phase). Figure 3 shows the observation and impact geometry. The observed impact rate during waxing phases is approximately 0.19/h and during waning phases is 0.07/h. During waxing phases the observed portion of the Moon is exposed to the antihelion, north and south toroidal, and apex sporadic sources while during the waning phases the observed portion is exposed only to the antihelion and toroidal sources. The apex meteoroids are impacting the lunar far side which we cannot observe. Clearly, the higher speed apex meteoroids at 55 km/s deposit much more kinetic energy than a similar sized antihelion or toroidal meteoroid at 25 km/s making their observed rate much higher than their flux would indicate. Thus, a clear signature of the apex source is present even in this relatively small data sample.

Fig. 3
figure 3

Observation and major sporadic source geometry. The observed impact rate is higher near first quarter because the earthshine portion of the moon (dark part) is exposed to the apex, toroidal and antihelion sources. At last quarter the rate is lower since only the antihelion and toroidal source meteoroids impact the observed portion of the Moon. Note that the north and south toroidal sources are out of the plane of the page

Full size image

The field of view of the camera encompasses approximately 10–12% of the total surface of the Moon. Assuming that helion meteoroids have the same flux as antihelion, this observed rate means that somewhere on the Moon, there are approximately 2–3 sporadic impacts per hour of sufficient energy to be observed from the Earth. These impacting meteoroids have masses of order 1 kg with a kinetic energy roughly equivalent to 200 kg of TNT. During meteor showers the rate increases dramatically, partially due to the flatter population index and hence larger percentage of larger particles.

4 Hypervelocity Impact Testing

In order to experimentally determine the luminous efficiency, a series of hypervelocity impact tests have been undertaken at NASA’s Ames Vertical Gun Range. Pyrex spheres of ¼ inch (6.3 mm) diameter were fired into ground pumice in a vacuum at speeds from 2.5 to 5.5 km/s and the impact flashes were recorded with the same StellaCam EX video cameras used for our lunar observations. Figure 4 shows the luminous efficiencies determined by the first series of shots in September 2006. The point in the upper right is the luminous efficiency η = 2 × 10−3 for Leonids determined by Rubio et al. (2000). All of the determinations of η have been plotted even though the impact angle was varied from 90° to 30° and the camera viewing angles varied between approximately 90° and 0°. Subsequent to these shots it was determined that the neutral density filters used to reduce the intensity of the impact flashes were not really neutral and had a factor of 10 higher transmissivity in the near infrared (where our cameras are sensitive and much of the thermal radiation from the impact is emitted) than in the visible. Thus the results in Fig. 4 are biased toward cooler impacts and should be treated as very preliminary. A second series of shots using truly neutral filters and lunar stimulant as the target material was completed in August 2007 and the analysis is underway. Future impact testing will be used to determine the mass flux, particle sizes, and particle velocities so that new cratering models can be validated and calibrated.

Fig. 4
figure 4

Preliminary luminous efficiencies determined from Ames Vertical Gun Range tests (left hand side of figure) and from Leonids observations by Rubio et al. (2000) (point at upper right of figure)

Full size image

5 Conclusions

NASA Marshall Space Flight Center has begun a campaign to observe sporadic and shower meteoroid impacts on the Moon. The fluxes of large impactors will be determined using luminous efficiencies from hypervelocity impact testing and shower meteoroid impact statistics. Further impact testing coupled with cratering models to be developed during this research will be used to calculate ejecta characteristics from impacts of various energies. The observed impact flux, sporadic source directionality from the Meteoroid Engineering Model (McNamara et al. 2004), and a Monte Carlo and orbit generation model will be used to propagate the ejecta around the Moon. This engineering model of the ejecta environment will be used by space hardware designers to build the systems needed to explore and establish permanent bases on the Moon.