The climatic signal in varved sediments from Lake C2, northern Ellesmere Island, Canada

Annually-laminated clastic sediments preserve a high resolution proxy record of paleoclimate, provided that allochthonous sedimentation represents a response to meteorological forcing of watershed sediment transfer. Here, we demonstrate this linkage, and illustrate a calibration process using the most recent 40 years of a varve record from Lake C2 (82~ t N; 78000 ' W), three years of field measurements, and meteorological data for 1951-92 from nearby AES weather station Alert. Field measurements were used to correlate proxies of the energy available for snowmelt (e.g. air temperature) and daily suspended sediment discharge (SSQ). Our calibration was extended through use of weather data from Alert. Both mean daily air temperature at Echo, and daily SSQ, were well correlated with air temperature at 600 m above Alert, as obtained from the 1200 Z (0800 LST) rawinsonde sounding. Accordingly, we used pooled 1990 and 1992 Alert 600 m data to predict the lagged daily sediment discharge into Lake C2 (adj. r 2 = 0.43). Daily values were summed each year in order to produce an annual series of predicted sediment transfer to the lake. The original varve chronology was based on eight sediment cores recovered from the deep basin of the lake (>80 m). Although low-frequency fluctuations of the varve and predicted SSQ series agree, slight tuning of the varve record optimizes the correlation between them. Adjustments were based on examination of weather data for specific years, reexamination of sediment core thin sections, and by aligning fluctuations in the two series which closely matched. Although the original chronology is reasonably well correlated with 600 m temperatures at Alert (for JJA mean, r= 0.41, significant at 0.01), the adjusted chronology is both better correlated and contains a more precise climate signal (r = 0.54 for July mean, significant at 0.01). This is the first calibrated varve record produced from Arctic lake sediments, and demonstrates that varves from Lake C2 contain a paleoclimatic record. We believe the post-facto manipulations required to produce the adjusted varve chronology are reasonable given the uncertainties inherent in varve counting, and the lack of any independent corroborating chronostratigraphic markers.