Catastrophic floods or drought? What caused the water level drop of glacial Lake Agassiz?

Lake Agassiz was an enormous freshwater lake formed by the melting and retreat of the Laurentide Ice Sheet (LIS).  The LIS, essentially a continental-sized glacier, covered a large portion of Northern North America with its southern extent spanning from Long Island, New York to Montana, approximately 20,000 years ago.  As the LIS melted, it retreated North and the meltwater collected into Lake Agassiz (Figure 1).  Lake Agassiz’s shorelines were dynamic through time, moving as ice-covered land became exposed.  As a result, the volume and areal extent of Lake Agassiz changed throughout its 5,000 year existence.  It has been estimated that Lake Agassiz covered an area as large as 841,000 km2.

Figure 1. A geographical reconstruction of glacial Lake Agassiz's total extent over ~5,000 years.  Arrows represent possible drainage routes and numbers represent the location of paleoclimate studies cited.
Figure 1. A geographical reconstruction of glacial Lake Agassiz’s total extent over ~5,000 years. Arrows represent possible drainage routes and numbers represent the location of paleoclimate studies cited.

Since Lake Agassiz no longer exists, scientists have relied on mapping geological environments associated with lakes, such as lake shorelines, to determine the area it once covered.  Sediment cores serve as additional tools for reconstructing a lake’s history.  As environments change over time (ie. swamp to shallow pond to large deep lake), different packages of sediment are deposited.  A sediment core can serve as a timeline for a particular location, documenting the changes in depositional environments.  Large drops in water level are relatively easy to recognize in a sediment core.  A particularly large drop in water level was dated to approximately 12,900 years before present (yr B.P.); a date shared by a drastic climate event called the Younger Dryas (YD).  This drop in water level, called the Moorhead low-water phase (Figure 2.) has been implicated in the YD climate event, a rapid return to cooler conditions lasting ~1000 years, during a time of gradual climatic warming.

Figure 2. Bathymetry of glacial Lake Agassiz prior to the Younger Dryas (A), at the Moorhead low-water phase (B), and after the water level increased following the end of the Younger Dryas (C).
Figure 2. Bathymetry of glacial Lake Agassiz prior to the Younger Dryas (A), at the Moorhead low-water phase (B), and after the water level increased following the end of the Younger Dryas (C).

Wallace Broecker, an already well-established oceanographer suggested in his 1989 paper that the onset of the YD was caused by a catastrophic release of freshwater from Lake Agassiz to the North Atlantic.  The North Atlantic is the location where cold dense water masses form and sink, largely driving global ocean circulation.  Broecker hypothesized that a large, rapid input of low density freshwater could disrupt the formation of North Atlantic Deep Water.  This would have a “domino-effect”, slowing ocean circulation and heat exchange to the northern latitudes, leading to climatic cooling (in this case, the Younger Dryas).  Broecker’s hypothesis has been widely accepted by the climate science community, though the routing of the freshwater to the North Atlantic remains an area of great debate.

Lake Agassiz had different drainage paths over it’s history.  Prior to 12,900 yr B.P., evidence in the Gulf of Mexico suggests a southward drainage of Lake Agassiz, through the Mississippi River. Each proposed drainage path has its own complications.  The drainage route to Lake Superior is tied to dates that do not match the beginning of the YD.  The Northwestern drainage route through the Mackenzie River system may have been blocked by the LIS.  The Eastward route through the St. Lawrence River system lacks high-energy drainage features such as boulders and gravel.  A competing hypothesis constructed from the inability to define an elegant drainage rerouting suggests the rapid drawdown around 12,900 yr B.P. was not a result of catastrophic drainage, but rather a change in Lake Agassiz’s hydrological budget.  In this paper, Teller evaluates all of the possible hypotheses and draws the most likely conclusion from decades of other researchers’ evidence.

The balance of water entering and leaving a lake is known as a hydrological budget.  For a lake to have a constant water level, the hydrological budget must be balanced.  The competing hypothesis suggests that Lake Agassiz entered a period where the hydrological budget went negative, drawing down the water level by as much as 50 – 100 meters.  This suggests a rapid change to dry conditions, where evaporation exceeds precipitation and meltwater input.  Teller, equipped with simple equations, solved for the rate of evaporation required to drawdown Lake Agassiz’s water level to the Moorhead low-level phase.   First, Teller provides the rates of evaporation required to offset the input of meltwater; numbers he estimated to be between 1.7 – 3.48 meters / year.  Teller put these numbers in perspective, reporting that the modern day Dead Sea basin (hot and very dry) experiences evaporation rates estimated at 1.1 – 1.2 meters / year.  Could Lake Agassiz have experienced evaporation rates exceeding the modern day Dead Sea basin?  To fully test this hypothesis, Teller would have to go one step further.

To reinforce his position that the evaporation rates needed to drawdown Lake Agassiz were unrealistic, Teller decided to look at the climate record during the Younger Dryas.  If indeed the climate record suggested conditions similar to the modern day Dead Sea basin, then perhaps the drop in water level 12,900 yr B.P. could be explained by an unbalanced water budget.  Teller cites eight different studies that detail the paleoecology and paleoclimatology of the Lake Agassiz basin and surrounding areas.  The numbers (1 – 8) on Figure 1. correspond to the location of each study.  None of the eight studies suggested a desert-like climate at any point during the YD.  The studies included descriptions of the plant and insect life found within the sediments.  The age of these organic matter can be estimated using radiocarbon dating techniques.  The age dated plant and animal life included species that would not survive in desert-like conditions.  The studies conclude that the YD near Lake Agassiz was actually a wet, cool period, with temperatures not far below modern day.  Combining the results of the eight studies with his hydrological budget calculations, Teller makes a strong argument against an evaporative Lake Agassiz during the YD.  Teller upholds that the drop in water level to the Moorhead low phase was caused by catastrophic drainage.

Understanding how our Earth’s climate system has changed in the past is essential to understanding how our modern climate will continue to respond to increased atmospheric levels of anthropogenic carbon dioxide. Uniformitarianism, or, processes that operate today are the same processes that operated in the past, drives scientists to understand events such as the Younger Dryas.  The Younger Dryas confirms that the Earth’s climate is capable of changing on very short geological timescales.  It has been nearly unanimously accepted within the scientific community that our modern day climate is rapidly warming.  The more scientists understand about our Earth’s climate system, the better prepared our societies will be to deal with future challenges that may arise due to climate change, with the assumption that there will soon be widespread acceptance of the science at a political and societal level.

Teller, J.T., Lake Agassiz in the Younger Dryas, Quaternary Research (2013),

One thought on “Catastrophic floods or drought? What caused the water level drop of glacial Lake Agassiz?

  1. Thanks for covering this issue. It seems Lowell et al. are somewhat reluctant to give this hypothesis up and have published yet another paper in support of it. The bigger fight is between the eastern drainage or the northwest drainage. And then there is the impact hypothesis.

    Little agreement between the major players it seems, which prompted me to propose an impact induced drainage to the east hypothesis, with a drainage remnant just south of Lake Nipigon in the embayment at the Black Sturgeon river basin.

    That hasn’t gone over very well.

Leave a Reply

Your email address will not be published.