Forming the NAVC—Correcting the NEVC

Errors in the NEVC

The reformulation of varve sequences in the northeastern U.S. as the NAVC involved the correction of Antevs� (1922, 1928) original �normal curves� for the NEVC and the joining of its two main sequences, the lower and upper Connecticut varve sequences. The corrections are based on new cores collected in the Connecticut Valley. The corrections include the addition of varves at various places that were missed by Antevs and also the joining of couplets where it appeared that there was an over count of varves. These errors are mostly in the older varves of the original NEVC (prior to NE 4500) where outcrops measured in the field by Antevs to compile these varves were usually wet, very clayey sections of thin varves. Under these circumstances thin varve couplets are sometimes difficult to see and this may have resulted in errors. Preparation of cores in a lab with careful partial drying and shaving of the cores has allowed a much better view of the varves than was available to Antevs. Still, very few errors have been found in the old NEVC with 23 new varves and 5 pairs of the original varves joined to form single varves over about 4400 varve years. Couplets measured in the Hudson Valley by De Geer (Antevs, 1922) and interpreted to represent single years with flood events are not single varves and have been corrected so that the 5 flood couplets now represent 1 giant varve (over 225 cm thick!!) with flood events in its summer layer (see discussion below). Altogether there are 14 more varves in the NAVC than in the original NEVC.

Errors and Uncertainty

Below is a table showing all of the corrections that were made to the old NEVC and how the NAVC incorporates a correction. All varves recognized as being different from most other varves, or just plain odd, are also listed on the table. Each entry also has a link to core images showing the spans of the corrections in new cores in the Connecticut Valley. The construction of the NAVC is based on the matching of multiple varve records, which is the only way to eliminate possible errors in most circumstances. The matches show a very high fidelity and there does not appear to be any uncertainty in the measurement or counting of varves except for those listed on the table. Only varves on the tail ends of the chronology are based on single outcrop records and have not been tested by matches to other sections. Uncertainties for the non-glacial part of the chronology are higher and individual uncertain varves (above AM 7282) are shown at NAVC Master Plots. A program and table for converting NEVC (NE) to NAVC (AM) years and vice versa are available for download at NAVC/NEVC Conversion. A set of master plots of the whole NAVC for every 100 years showing where corrections were made has also been assembled and is given at NAVC Master Plots.

Correcting the Hudson Valley "Flood Varves"

While formulating the NEVC, Antevs (1922) could not find sections in the Connecticut Valley that spanned an interval from NE 5601-5708. Antevs used sections in the Hudson Valley measured by Gerard De Geer (reported by Antevs, 1922) that spanned 300 years, matched NE 5501-5600, spanned the gap, and matched NE 5709-5800 thus plugging the hole in the Connecticut Valley. In the gap the NEVC Hudson Valley record had five consecutive extremely thick couplets (up to 60.8 cm) in one interval that De Geer interpreted as individual varves deposited during successive years of flood events. A long core of varves on the campus of the University of Massachusetts at Amherst (UMass core) provided the first look at the interval of the gap in varves from the Connecticut Valley. This core was studied by Tammy Rittenour (1999) as part of her MS thesis under the direction of Julie Brigham Grette in the Department of Geosciences at the Univ. of Massachusetts. Tammy discovered that the years of the large flood “varves” in the Hudson Valley were missing from the Connecticut Valley record. This provided a test of the annual nature of the flood couplets and showed that they were really all events in the summer layer of one very thick varve!! This single varve with 5 flood events has an annual thickness of just over 225 cm and represents the drainage of a lake in the Mohawk Valley (stage of Lake Amsterdam) into the Hudson Valley during one summer. Shown below is a plot of varves from the UMass core and the old NEVC as compiled in Rittenour (1999).

Caption: Plot of the UMass core with its correlation (red; Rittenour, 1999) to the Hudson Valley record of De Geer (magenta) as reported in Antevs (1922).  Also shown is a plot from Canoe Brook, Dummerston, VT (black). Varves NE 5669-5678 do not appear in the UMass core as correlated by Rittenour with no apparent erosion at the UMass site. On the plot, varves NE 5620-5667 match very well while varves immediately after the flood events in the Hudson Valley record (NE 5679-5720) are noticeably thicker than Hudson Valley varves prior to the flood and do not seem to match the UMass core record very well. Click to see larger image.

The importance of the UMass core is that it not only demonstrated where a correction needed to be made in the NEVC but it establishes the timing and character of the drainage of a Mohawk Valley glacial lake into the Hudson Valley. The drainage event occurred over the course of just one summer either as distinct successive falling water level stages or perhaps with partial refilling prior to each later drainage event.

Matching of a new core from North Hatfield, Massachusetts (NHT core) to the NEVC revealed that the interval of the flood events had some other NEVC errors that needed to be addressed. While correcting these errors the match of the UMass core has been revised from what is shown by Rittenour (1999) with the match from NE 5680-5725 being interpreted differently. Below is a plot of the new interpretation of the NAVC (corrected NEVC) chronology and its correlation to the UMass and NHT cores.

Caption: Plot of NAVC (corrected NEVC) varves in the Hudson Valley (magenta), NHT core (green), corrected UMass core (red), and Canoe Brook, VT record (black) for AM 5620-5720. Note that on the NAVC time scale here the Hudson flood events are offset by about 15 years from the NEVC plot (above) because of corrections lower in the NAVC that have added years to the chronology numbering system. Click to see larger image

Closure of the Claremont Gap

The discovery of the true overlap of the lower and upper Connecticut varve sequences of the NEVC (NE varve years) occurred during the measurement of new cores at the Aldrich Brook site in Westmoreland, NH and in the Perry Hill Basin in Charlestown, NH. In both places preserved lake floor surfaces were cored and yielded long varve records that connect the two old NEVC sequences. This has allowed the lower and upper Connecticut varves of Antevs� NEVC to take on the same numbering system and a single continuous sequence has been constructed using the NAVC (AM varve years) numbering system. For more on how the Claremont Gap was closed in 2008 go to NEW ENGLAND VARVE CHRONOLOGY/Claremont Gap Closure

References

• Antevs, E., 1922, The Recession of the Last Ice Sheet in New England: American Geographical Society Research Series, no. 11, 120 p.
• Antevs, E., 1928, The Last Glaciation with Special Reference to the Ice Sheet in North America: American Geographical Society Research Series, no. 17, 292 p.
• Rittenour, T.M., 1999, Drainage history of glacial Lake Hitchcock, northeastern USA: Unpublished M.S. Thesis, Univ. of Massachusetts, Amherst, 179 p.

2012 Geolimnology Newsletter Figures

Click Here!

Varve Project News