Searching for the Sands of Time

Each of the cores I collected in June brings to light new information about historical changes in sediment input to my field site from the Merrimack River. Each grain of sediment began its journey somewhere in the extensive Merrimack River watershed, from the rugged, rocky White Mountains of New Hampshire to the eroding, glacially-carved landscapes of the Merrimack Valley and northern Massachusetts. Together, the grains begin to form distinct layers, based on their size, shape, mineral composition, and organic content, among other characteristics. These characteristics track changes discernible by sight and touch, old but useful tools of sedimentology.

Recently I used a cutting-edge tool of sedimentologists and other scientists: an XRF Core Scanner. The scanner uses x-ray fluorescence to provide high-resolution (centimeter-scale for our purposes) analysis of the elements present in a sample of sediment. The machine I used at the University of North Carolina (Chapel Hill), an Avaatech XRF Core Scanner, can scan a core up to a 1.5m long in one run. Each run can take anywhere from 45 minutes to 6+ hours depending on the length of the core and the instrument settings. While 6 hours may seem a long time, it is extremely quick compared to other laboratory methods, especially given the scanner’s high-resolution and the amount of data collected.


The X-Ray Fluorescence core scanner at UNC Chapel Hill provides high-resolution elemental analysis of sediment cores

The machine captured changes in a variety of elements, but the ones most important to this project were industrial contaminants such as arsenic and zinc. The presence of a high concentration of these elements in a particular layer of sediment captures the peak of the Industrial Revolution. Not only does the layer provide evidence of how humans can contaminate sediment by influencing its chemical composition, but the layer also provides a timestamp or date somewhere between 1850 and 1920. Timestamps throughout the length of a core are a key piece of answering the questions about how sediment input  has changed over time. For example, we want to know if the rate at which sediment was deposited between 1850 and 1950 is different than the rate between 1950 and today.

While today the industrial presence in the Merrimack River valley is minimal, the Merrimack was a key piece of New England’s industrial prowess in the 19th Century. Indeed, the famous textile mills at Lowell were located along the Merrimack River. Textile manufacturers and other industrial barons diverted and dammed water from the river to power mills, trapping large volumes of sediment behind dams. As well, contaminated effluent from dyeing and other industrial processes was dumped directly into the river where it contaminated the sediments in and around the river to varying degrees. John T. Cumbler describes the industrial practices and history along the Merrimack and other New England rivers quite well in his book “Reasonable Use”—well worth a read.

Besides reading up on industrial history, I am cleaning up and combing through the treasure-trove of data from the scanner. In the coming weeks, I’ll have a better sense of what the data show and how they help answer the central questions of my research. The scanner data are a key component of the success of this research and would not exist without the generosity of Dr. Laura Moore and Anna Jalowska at UNC Chapel Hill!


  1. I would be interested to know more about the microscopic and macroscopic organisms you found in your sediment samples, though I know this was not the primary focus of your project. Did you see a decrease in organisms/microorganisms in the more polluted layers of sediment? I suppose any difference in biological communities at this level would have to be between anarobic species, as I would think that the layers of sediment formed during the industrial revolution would be too deep to gain oxygen. Still, could you tell me about any biological trends you noticed?

  2. This machine seems fascinating. You talked about the resolution of the scanner, which in your case is down to centimeter level, but what about the precision and density of the time stamps you establish? In other words, how many dates can you pull out from spikes of certain elements throughout the core and how confident are you about those points? You also mentioned that damming was an important anthropogenic factor in sediment deposition. Could that practice have any lasting bearing on the size and distribution of tidal flats downstream?

  3. This is so cool!