Projects by Edward Earl
Star Finder, 1983-1984.
This is a planetarium simulator for the Apple ][+ and Apple ][e.
It was a commercial product, available from Earl Enterprises.
ESR Spectrum Simulator, 1984-1986.
This package was used in research to help interpret and analyze ESR
spectra. It proved to be a very valuable scientific tool. Many scientific
results could be understood and explained with the aid of this package,
which can handle very general cases robustly.
Digital Tachometer, 1988.
This was a hardware project in which I designed and built a digital
tachometer for my VW bus using parts from Radio Shack and other electronics
suppliers. The design included several analog and digital chips and two LED
digits which showed engine RPM in hundreds.
Major contributor to the Utah Mess Kit, 1987-1992.
The Utah Mess Kit is a quantum-mechanical electronic
structure package. Projects that I undertook therein include:
- Integral transformations
- High-symmetry adaptations
- Graphical and menu-driven user interfaces
- Direct CI calculations
Symbolic Manipulator, 1993.
This was unique in that it allowed applications to define their own
kinds of symbolic objects with their own algebraic properties. Only a modest
quantity of application callback code would enable this package to perform
group theory and abstract algebra. If the project had not been discontinued
due to a restructuring of the company, this package would have
substantially increased the flexibility of BIOSYM's Discover
molecular simulation package.
Altimeter calibration, 1993.
This was a personal project to measure the parameters of a digital
electronic altimeter. An airplane flight was performed to collect readout
data as well as the readings on the aircraft altimeter. A nonlinear
least-squares fit was used to calculate the atmospheric conditions assumed
by the electronic altimeter, along with their uncertainties as determined by
the amount of scatter in the data. Correlations between the uncertainties of
the different parameters were also calculated. The altimeter was found to be
calibrated to a standard atmosphere within the error of measurement.
Solar Eclipse Timing, 1994.
This methodology used data from the Astronomical Almanac to
predict the time of a solar eclipse at any location. The methods were
designed so that using some precalculated data, one could venture to any
eclipse site with only a hand calculator (no computer necessary) and
determine the exact time of the eclipse, taking the effect of elevation into
account. The method was used successfully to predict the eclipse on
November 3, 1994, as seen by me from Sevaruyo, Bolivia, within one second.
Topographic Prominence, 1998-present
In mountain climbing circles, "prominence" refers to the height of a
peak above the highest saddle connecting it to a higher peak. This effort
automated the process of calculating the prominence of every peak in any
area, given digital elevation data for that area. That task is very tedious
and time-consuming if done by hand. Scaled up, the implementation can
analyze the entire United States in only a few hours of processor time.
The algorithm is so clever and novel that it may be patentable. Today the
software that performs this task is a professional grade Windows MFC GUI.
Photo Pixel Analysis, 2008
The objective of this project was to determine which of two rival peaks is
higher, given a photograph taken from one peak looking toward the other, or
of both peaks from a nearby vantage point. The photograph should include
several other peaks of known height and position for calibration of the image.
The pixel coordinates of each calibration peak and rival peak are provided to
a statistical analysis that determines the height of each rival peak relative
to the point from which the picture was taken. The method was successfully
applied to Buckner Mountain, where it was determined that the southwest
summit is 10 inches higher than the northeast summit, with an error of plus or
minus 4 inches. The method can also be used in "forensic" analysis to
determine where a picture was taken from.
Beta Distributions, 2009
A software tool computes the distribution of a sum, difference, maximum,
minimum, conditional maximum, and conditional minimum of multiple beta
distributions over arbitrary intervals. Although beta distributions are
essentially polynomials, serious numerical errors result from manipulating
the distributions as explicit polynomials. Therefore, considerable effort was
devoted to manipulating the distributions as linear combinations of beta
distributions instead. The application of the project was to calculate the
probabilities that two or more mountain peaks have a given relative prominence,
where prominence is defined as in the "Topographic Prominence" project
described above.
Back to my
main resume