Daniel J. Breton

Current work


Much of my Ph.D. work has focused on designing, building, and testing an instrument called MADGE: the Maine Automated Density Gauge Experiment. This is an instrument designed to be the best density gauge possible for the materials that the Climate Change Institute primarily works on: 5 to 8 cm diameter firn and ice cores. I had several major goals for the instrument design:
  • Uncertainty in density < ±0.005 g/cm3
  • Vertical resolution of 3.3 mm
  • Portable/tough enough to be usable in the field on an Antarctic traverse
  • Throughput > 1 m/h
  • Able to operate and store data independently of a PC
The instrument that resulted has a sensor head (containing the source, detector and electronic calipers to perform the density measurement) mounted on a linear actuator. This actuator is driven by a stepper motor which provides very fine control over the sensor head position as it moves along the firn/ice core. Control of the stepper motor driver, nuclear counting electronics, various temperature control circuits, data storage and user interface are all performed by a single Rabbit RCM3720 microcontroller which is much better suited to deal with -40 C than a typical laptop, and is significantly cheaper.

When field work began on ITASE 2006, MADGE was able to perform better than the design goals, and of course more improvements are in the works, so our current 'normal' performance level of 1.5 m/h at ±0.004 g/cm3 for a nominal 0.5 g/cm3, 5.1 cm diameter core may improve over time. We have a lot of confidence in our calibration technique and have also compared density gauge data against manual measurements in the field with excellent results. Modifications to the sensor head will soon allow us to profile 8 cm cores which should minimize some of the problems of core quality that we experienced with the smaller, more fragile 5 cm cores.


While conducting field work with MADGE on 5cm firn cores in East Antarctica, it became clear that significant amounts of firn were not recovered (i.e. destroyed) in the coring drill due to the extreme weakness of buried hoar layers and the small diameter of the core itself. I wanted a way to quantify the depth and extent of the non-recovered core sections, so I developed MABLE: the Mostly Automated Borehole Logging Experiment. This is an instrument that logs borehole wall hardness and infrared reflectivity to determine where core loss is most likely.

Initial work with combined MADGE and MABLE data from a Titan Dome core indicate that we can track individual density-hardness and density-reflectivity events to get a good picture of how much core was lost and where. This information allows us to put data gaps at proper locations in the density profile so that the density reported at a given depth is the true depth, not just the length down the core.

Many people ask, "Why only mostly automated?" I had a limited amount of time to design and build before leaving for Antarctica, and at the time it was obviously much simpler and cheaper to have graduate student (me) to carefully operate a manual winch than have some sort of complicated motorized thing. All of the data collection is automated, but the the process of lowering and raising the instrument in the borehole is not.

Also, I am a sucker for a decent sounding acronym. Both instruments are still prototypes and therefore the final E in both acronyms usually stands for Experiment, but has taken on different meanings from time to time depending on the temperament of the instrument at that moment:
  • Experience
  • Explosion
  • Exasperation
  • Error
  • and of course
  • Expletive
Links to more detailed information about the instruments

Climate Change Institute 2007 Symposium presentation [pdf]
A presentation of calibration and some actual data after the first season of Antarctic field work. At this date, MABLE was merely a twinkle in my eye, so to speak... 2.1MB.

Dartmouth Firn Workshop 2008 presentation [pdf]
This presentation contains some general information about both MADGE and MABLE and also shows some data collected from each... 3.8MB.

Poster on Gamma-ray Density Gauge Design [pdf]
This poster gives (I hope) a decent overview of the design considerations/constraints involved in designing any density gauge... 1.8MB.

Hopefully this list will soon include an actual peer-reviewed article, as well.