Reconstructing hydrological events from annually laminated sediment in northwestern British Columbia

Determining what has happened in a landscape in the past is a key sub-discipline in geomorphology – through various techniques, aspects of these past environments are reconstructed. For example, examining glacial sediment deposits on a landscape (e.g., moraines, eskers, drumlins) can help researchers to reconstruct where the glacier was at one time in the past. We can use sediment deposits in the bottom of a lake, as a record of sediment transport and delivery (via rivers) in that watershed. If we want to understand past river flows and rates of sedimentary processes, sedimentary records from lakes are excellent archives of these environmental variables. Past environmental conditions are useful to understand because they help us to see rates of change, magnitudes of variability and may give insight into how these environments may respond to future changes.

In this vignette, we will examine the recent (last ~700 years) sedimentary record from Summit Lake in northwestern British Columbia, Canada to evaluate past runoff in the catchment (Figure 1). Hydrological events (runoff events) are defined as episodes of increased runoff due to specific forcing mechanisms. For example, spring runoff in the Summit Lake watershed is driven by snowmelt, the sediment deposited in Summit Lake in the early part of the year reflects this process. As snowmelt subsides, summer glacial runoff (from the small cirque glaciers in the headwaters) would increase and the subsequent sediment transferred and deposited at that time reflects this process. As glacial melt is highly dependent on summer temperatures, we would expect glacial melt to be variable over the course of a single summer. When there are a few warm days in a row we would expect higher sediment yields and when there are a few cool days in a row we would expect that glacial melt and thus river flow would subside, and result in a change in the amount and type of sediment delivered to the bottom of Summit Lake. In this part of British Columbia, the late summer and early fall (before precipitation is dominated by snow), short-lived, but intense rainfall can occur. And similar to snowmelt and glacialmelt runoff, we expect to see sedimentary evidence of increased stream flow due to rainfall.

Summit Lake is a good place to study various hydrological events because sediment is seasonally delivered to the lake and once deposited it is not disturbed by biota or other activity (wind, boats, etc.). As well, the high availability of sediment, specifically fine grained (clay-sized) material, means that sediment supply is not limited. The fine-grained material settles to the bottom of the lake and accumulates in a concentrated unit known as a clay-cap, over the winter which is overlain by the material delivered to the lake in the spring via snow melt, which is typically larger in size (silt to sand size). This rhythmic delivery and deposition of sediment to the lake bottom produces laminated sediments; furthermore we can confirm that these laminae are annual (one combination of layers equals one year). This is sometimes called a varve. Thus we can reconstruct hydrological events (based on the amount of sediment and its grain size) and assign a calendar year to the deposit (Figure 2).

These hydrological events generate sedimentary deposits in the lake. Following stratigraphic position within the annual sedimentary unit (fine clay cap, consistent thickness each year, open rectangle in Figure 2) hydrological events are counted and measured. The total thickness (reported in Figure 3) is a metric for the amount of sediment delivered, and the number of events in each year (reported in Figure 4) is a metric for the variability in the delivery of sediment to the lake. During the spring melt season and throughout the summer glacial melt period sediment transported to and deposited in the lake reflects the sediment available and runoff energy. Late summer and early fall rain storms generated by the seasonal low that develops off the coast. The late summer/early fall rain events often interrupt the clay cap at the top of each annual unit and are frequently coarse grained with small pieces of organic detritus.

Both varve complexity (number of events) and the timing of the events increased at the end of the 17^th century (1670s). This is coincident with increased Pacific Decadal Variability (PDV) estimated through a number of other paleoclimate proxies (archives of past climate). Varves are the proxies discussed here. Tree-rings, isotope ratios, pollen fossils are examples of others. PDV is known to influence the tracks of low-pressure systems over the Gulf of Alaska, which would have increased rainfall events at the end of the summer and likely increased the snowfall and potentially influenced summer melt events. We would expect to see an increase in the number of events (varve complexity) in a given year (Figure 4) and potentially thicker annual units (Figure 3). Reconstructing past events helps researchers to understand how climate and geomorphic systems functioned, which allows us to make predictions of how these systems may respond to future variations in conditions.

• Cockburn, JMH and Lamoureux, SF. 2007. Century-scale variability in late-summer rainfall events recorded over seven centuries in subannually laminated lacustrine sediments, White Pass, British Columbia. Quaternary Science Reviews, 167: 193-2003.