Replication of “An Experiment to Measure the Absolute Motion of the Earth,” Part 2

Lance Osadchey, C. Michael Edwards, and Pat Kirol

Dr. Lance Osadchey's article "An Experiment to Measure The Absolute Motion of the Earth" (The Citizen Scientist, 02 February 2007) was followed by various letters in "Backscatter," a major article by C. Michael Edwards ("A Progress Report on Replicating An Experiment to Measure the Absolute Motion of the Earth," 6 July 2007) and the record number of views (8,092 as of this writing) for any topic on the Society for Amateur Scientists Forum.

Observations and Results

The observed deflection of the laser beam manifests as a difference in the mean position of the laser central peak at a horizontal instrument orientation angle, a, versus the mean position at the angle a ±180°. This displacement varies sinusoidally over each transit of the instrument with a period of 360°. The displacement has a vertical and horizontal component, resulting in oscillatory motion in the photograph reference frame.

Figure 8. Mean image positions for Dr. Osadchey's 11/05/07 experimental trial, with fitted curve for motion.

Figure 8 shows computed motion for Dr. Osadchey's trial conducted at 44.0°N 72.1°W, beginning at 20:15 CST on 5 November 2007. n = 4 turns with 30º intervals (a_initial = a_final = 1440º, and delta a_step = 30º). Instrument elevation is 0º above horizontal. Position has been adjusted for drift. The estimated pixel dimension is 4 ± 0.5 µm square, based on manufacturer's information. The component amplitudes of the fitted displacement curve shown are 4.1 pixels horizontal (s = 2.4 pixels) and 8.9 pixels vertical (s = 3.2 pixels). This corresponds to component amplitudes of 16.3 ± 9.7 µm horizontal and 35.6 ± 12.7 µm vertical. The motion is oscillatory but not circular. The maxima of the fitted component curves occur at a_x = 17º ± 30º for horizontal and a_y = 199º ± 30º for vertical.

Figure 9. Mean image positions for Mr. Edwards's 11-05-07 trial, with fitted curve for motion. This trial was conducted simultaneously with the trial shown in Fig. 2.

Figure 9 shows the computed mean position results for the simultaneous trial conducted at 31.3°N 92.4°W, beginning at 20:15:17 CST on November 5, 2007 with the same procedure as the trial represented in Fig. 8. The pixel dimensions are 2.9 ± 0.3 µm/pixel horizontal and 3.2 ± 0.3 µm/pixel vertical based on CCD chip dimensions and camera resolution. The component amplitudes of the fitted displacement curve shown are 1.8 pixels horizontal (s = 1.5 pixels) and 1.9 pixels vertical (s = 2.2 pixels). This corresponds to component amplitudes of 5.2 ± 3.3 µm horizontal and 6.0 ± 7.0 µm vertical. The maxima of the fitted component curves occur at a_x = 48º ± 30º and a_y = 94º ± 30º. The vertical data for this trial fail the acceptance criterion Y > s (the vertical error bars all overlap), but the horizontal data is usable.

The trials plotted in Figs. 2 and 3 were conducted with simultaneous start times as an experimental control. The fitted amplitudes vary with start time for both locations, but the ratio of horizontal amplitudes shown is typical. The two stations do not measure the same amplitude, with maximum displacement at 44.0°N 72.1°W greater than at31.3°N 92.4°W. However, after applying a correction of 20.3º to account for the difference in longitude, the angles of maximum horizontal displacement at both stations agree to within experimental accuracy. This relationship is typical.

The stations are 2,200 km apart, and both signals have approximately the same horizontal phase.

Models of the causative agent based on known causes of misalignment were evaluated, but proved inadequate to explain all of the signal characteristics, including: measurement interval independence, lab frame position independence (including rotated and inverted positions), geocentric frame position dependence, and lack of variation with controlled changes in instrument structure.

We got an interesting result by modeling the causative agent as a hypothetical vector field creating light aberration. A vector product description was selected as an empirical model. The model outcome yields no displacement component along the light path when controlled for lateral displacement, and its vector geometry matches the instrument geometry closely enough to be tested directly against experimental results.

where delta s is the vector difference in displacement (i.e., the deflection) along a path of displacement s_o without aberration, and A is the mean vector of the causative field. Variation in the displacement per unit length is neglected and A is assumed constant. The components of the light path displacement vector in the lab reference frame are:

where L is the path length without aberration, b is the instrument elevation above horizontal and a is the instrument azimuth angle on a cardinal degree scale. (Note: a_cardinal = -a_polar.) Similarly, the components of the causative field in the lab's reference frame are:

where a_A and b_A are the azimuth cardinal angle and elevation of the causative field.

Equivalent coordinates can be defined in the instrument's reference frame.

Figure 10. The instrument reference frame.

where A' and s' are the causative vector and the light path displacement without aberration as expressed in the instrument's reference frame, respectively, and X and Y are the measured components of displacement in the surface plane of the camera photodiode. The rotation of the instrument during measurement is about the vertical axis with constant b, and adjustment of b is accomplished by rotation about the x' axis. Thus, the measured components of delta s are:

These are the only components of delta s in the instrument reference frame.

The maxima of these functions are:

The causative field azimuth can be determined from the angle of maximum X. b must be held constant to reduce deformation of the instrument support beam. The equations for X and Y can be solved for the elevation of the causative field.

Figure 11. A plot of computed angles of horizontal maxima vs. sidereal time.

Figure 11 is a graph of horizontal phase vs. sidereal time. It shows no direct dependence on sidereal time. This data does suggest a time dependence, however. Two groups of data points recorded at 23 ± 1 hrs sidereal time are bracketed. The group about 0º and 360º was recorded in June 2007 and July 2007. The group about 180º was recorded in December 2008 and January 2008. We propose that the 180º separation between the medians of these clusters is not an artifact, but corresponds to the angle swept by the solar zenith in heliocentric space during the intervening time.

Figure 12. A graph of the computed angles of horizontal maxima vs. solar time for the data in Fig. 11. A polynomial fit is applied to provide an estimated maximum.

Figure 12 shows that the horizontal phase is dependent on solar time, with a maximum at approximately 06:00 local time. The existence of an inflection point in the fitted curve indicates that the causative field reverses direction in the lab frame over the course of one day. The maximum precedes the sun's zenith by approximately 6 hours.

Figure 13. Cardinal angle of horizontal maxima vs. angle from sun in lab frame. A linear fit and polynomial fit have been applied to illustrate inflection.

Figure 13 shows the relationship between horizontal phase angle and the corresponding angle from the sun of the horizontal maximum in the instrument frame. The fitted curve corresponds to daily variation of the sun's altitude in the instrument reference frame at the lab's geographic location. Data was collected over the course of a year, and shows no regular variation with measurement date or temperature.

Figure 14. Fitted curves for adjusted image position. Corresponding CST measurement times are listed for each trial.

Figure 14 shows fitted curves for adjusted image position computed for measurements taken at various solar times. The daily phase variation in these curves corresponds to that shown in Fig. 12. Daily amplitude variation is apparent as well. There are two image axes for which the aspect of the adjusted image position curve reduces to that of a line and the apparent aberration is planar rather than elliptical, as illustrated by data taken at 13:27 and 21:40. These correspond to orientations for which the vertical and horizontal components of the displacement in the instrument frame are approximately in constant proportion throughout the duration of the trial.

Reversal of the vertical component over 360° is observed in all trials meeting statistical acceptance criteria at all stations.

Figure 15. Adjusted horizontal coordinates for two lasers operated simultaneously.

Simultaneous operation of a second instrument as a control is shown in Fig. 15. The two systems used the same mount with opposite directions. The magnitudes approximately agree, and the computed phase difference between the two is 0°.

Grounded shielding and filter circuits to counteract EMI tended to reduce image noise but not amplitude. These measures refined the signal, without reducing or eliminating it. EMI countermeasures produced no change in the phase vs. time relationship.

Experimentation with a wheeled platform rather than a fixed mount allowed us to control for variations in the thermal environment, as well as inclination and/or defects in the lab floor or mount. Use of a magnetic compass eliminated reliance on external markers for orientation, allowing measurements at the same orientation without systematically repeating the same location. The experiment was conducted in multiple locations with position varied between measurements over the course of each trial. No position dependent variation was observed.

Lance Osadchey, Pat Kirol and Hayden Brownell have performed long duration stationary trials to test daily variation in stationary instruments. Osadchey and Brownell both obtained results indicating solar time dependence. Pat Kirol reported an indeterminate result with a design using a reflecting mirror. However, a control experiment performed by Dr. Osadchey indicated that two simultaneously conducted stationary trials can produce different readings due to drift, and Pat Kirol's results clearly show drift of the same magnitude as the measurements by Osadchey and Brownell with a period less than four hours. No stationary trials conducted to date have a means for drift correction.

Conclusions

Consistency of results regardless of changes in instrument design and structure, lack of variation with location changes, phase alignment between stations on a geographic scale, and comparison with induced systematic errors indicates a causative agent external to the instrument.

The vertical phase does vary with sidereal time. However, the horizontal phase does not show any such relationship, varying only with local time. The variation is not temperature dependent; however, when using an empirical model of light aberration due to a vector product interaction, the direction of the hypothetical vector field matches the astronomical position of the sun. The sidereal variation of the vertical phase is consistent with declination changes of the sun.

Dr. Osadchey's Velador is tracking the sun.

This result refutes Dr. Osadchey's earlier claim that this instrument is measuring an absolute reference. If the hypothetical causal vector field is central, with an origin inside the solar system, then it can't be absolute. The empirical model presented here has arbitrary geometric constraints for which we are still investigating alternatives, but a different geometry won't negate time dependence.

This result, incidentally, contradicts the luminiferous aether model of light propagation. Aether does not flow from the sun in classical models, or have a vector product interaction with light. Further, the vertical oscillation in the absence of vertical rotation also contradicts the predicted behavior of light in a luminiferous aether flow field. There should be no well-defined vertical component observable in such a field.

This outcome is a novel disproof of the luminiferous aether flow field model for this phenomenon.

However, we are still seeing a regular, orientation dependent deflection that we cannot account for. We have tested several models for systematic errors proposed by other researchers. To date, we have found these models inadequate to predict the behavior. (Similar deflections have been reported as systematic errors in interferometers with unstabilized sources dating back to Michaelson and Morley in the 19th century. The deflection is also on the same scale as reported orientation dependent misalignments in interferometers with stabilized sources, such as cube-corner retroreflector defects. Whatever else it may be, we are confident that this is an old friend of interferometry.)

The luminiferous aether flow model is wrong, but the conventional explanations we tested are demonstrably wrong as well.

For example, varying temperature, soak time, and thermal insulation indicates that the oscillation amplitude is independent of both heat flow rate and net heat flux at the instrument's outer surface, as well as drift rate. Also, reduction of the drift rate per turn to less than the computed amplitude where that amplitude remains relatively constant over long periods (trial durations up to 2 hours) suggests that the system entropy is not increasing with each oscillation. Producing this behavior with an external heat source would violate conservation of energy and the second law of thermodynamics.

That does not preclude an internal heat source (for which we can only control soak time, not heat flow rate). However, an internal heat source requires some other restoring mechanism to oscillate. That mechanism would need to be orientation dependent.

Our observations also constrain the orientation dependence, further limiting the possible causes. Because the horizontal phase regularly varies 360º per day, and the effect is observed regardless of equipment parameters, it cannot be attributed to AC current driven interference. Countermeasures alone cannot eliminate all interference. Thus, we cannot discount EMI as a cause using our results to date. To fully account for EMI, an assessment versus geographic location is required. EMI due to human activity will not be constant or regularly variable on geographic scales. Our current results are not constant, either, but would vary sinusoidally with latitude if our assumption of a constant vector field holds true on a geographic scale.

Our results continue to suggest an astronomical phenomenon, but we lack a crucial measurement to prove that assertion. We suspect that the differences in our results correspond not to instrument parameters (our instrumentation was exchanged several times for equipment with different electrical and thermal parameters as well as swapped by mail between stations and continued to yield the same result) but to vector field changes with latitude. If that is so, then there should be predictable consequences of variation over time zones and latitude. To show latitude dependence, we need to obtain a wider range of data at different geographic coordinates.

Toward this end, we are continuing our experiment.

References and Recommended Reading

1. Devore, J.L. Probability and Statistics for Engineers and the Sciences , Third Edition. Brooks/Cole Publishing Co., c. 1991.

•  Osadchey, L. “An Experiment to Measure the Absolute Motion of the Earth,” The Citizen Scientist , 2 Feb. 2007, URL http://www.sas.org/tcs/weeklyIssues_2007/2007-03-03/project1/index.html , Last Accessed 5/22/08.

•  Osadchey, L. “Update and Comments about An Experiment to Measure the Absolute Motion of the Earth,” The Citizen Scientist, 6 Apr, 2007, URL http://www.sas.org/tcs/weeklyIssues_2007/2007-04-06/project2/index.html , Last Accessed 5/22/08.

•  Osadchey, L. “The Velador Series”, URL http://www.lanceosadchey.com/ , Last Accessed 5/22/07.

•  Roberts, Thomas J. “An Explanation of Dayton Miller's Anomalous ‘Ether Drift' Result.” URL http://arxiv.org/ftp/physics/papers/0608/0608238.pdf , Last Accessed 5/22/08.

•  Edwards, C.M. “Reproducing the Velador Experiment.” URL http://hearth2.50megs.com/ , Last Accessed 5/22/08.

•  Edwards, C.M. “A Progress Report on Replicating an Experiment to Measure the Absolute Motion of the Earth.” The Citizen Scientist, 6 July 2007, URL http://www.sas.org/tcs/weeklyIssues_2007/2007-07-06/project1/index.html , Last Accessed 5/22/08.

•  Lienhard, J.H., IV and Lienhard, J.H. V. A Heat Transfer Textbook , 3 rd Edition. Phlogiston Press, c. 2005. URL http://web.mit.edu/lienhard/www/ahtt.html (Public Access - Registration Required), Last Accessed 5/22/08.

•  Blocki, J., et. al. “Laser Measurement of the LumiCal Detector Displacement.” IFJ PAN, Report No. 1985/PH, December 2006. URL http://www.ifj.edu.pl/publ/reports/2006/1985.pdf?lang=pl , Last Accessed 5/22/08.

•  Roberts, Thomas J. “What is the Experimental Basis of Special Relativity?” URL http://math.ucr.edu/home/baez/physics/Relativity/SR/experiments.html , Last Accessed 5/27/08.

Readers are encouraged to send comments about this article to "Backscatter" at editor[at]sas[dot]org.