Skip to main content

Chilean Lake sediments as archives for climate variability during the past 1000 years

Final Report Summary - CHILE1000 (Chilean Lake sediments as archives for climate variability during the past 1000 years)


The main aim of the research project CHILE1000 (project no.276703) was to develop quantitative summer and winter temperature reconstructions for the Chilean Andes covering the past 1000 years and place these in a wider (spatial, temporal and climatological) context. This should be achieved by studying the composition and characteristics of lake sediment cores retrieved from small, remote lakes at high altitudes. So far, few paleolimnological studies with the particular aim to reconstruct temperatures quantitatively have been carried out in this region. Therefore, to achieve the main aim in this project, a number of relatively new techniques needed to be developed and tested. These are presented in more detail below.

In summary, the research project was successful in achieving a number of important aims:
We produced the first, quantitative, high-quality training set for chrysophyte stomatocysts in South America. During an extensive and highly successful field campaign in summer 2011, thirty remote lakes in the Chilean Andes were equipped with sediment traps and thermistors which formed the basis for the training set. All planned materials (most importantly: short and long sediment cores, water samples) and data (in situ measured water conductivity, alkalinity, depth, clarity and catchment characteristics) were collected as planned. During a second field campaign, traps and thermistors were found and materials and data were collected in 24 lakes. From the sediment trap material, chrysophyte stomatocysts were concentrated and these were analysed in detail. In addition, water chemistry was analysed, thermistor (measured temperatures of the surface waters) and data were extracted and analysed. All put together, these thorough analyses resulted in 47 carefully measured and ecologically relevant environmental parameters which characterized each of the 24 lakes in detail. Next, the modern chrysophyte stomatocyst assemblages were ‘compared’ to the environmental parameters using advanced statistical analyses. A statistically significant and robust Transfer Function was developed, quantifying the relation between stomatocyst assemblages and winter-spring temperatures. This finding is highly relevant, since so far the sensitivity to cold season temperatures was only shown conclusively for the Austrian and Swiss Alpine region and the Spanish Pyrenees. We now successfully applied this novel methodology to a completely new region, and the results are in excellent agreement with the work from Europe. The proxy is now firmly established as a highly reliable indicator for past winter conditions, which makes it unique since nearly all other natural proxies reflect summertime conditions.

Using this methodology, we then produced a reconstruction of past winter-spring conditions back to AD 160 for Laguna Escondida. This is a highly remote lake situated on the western flanks of the Andes at 38°S (1740 m above sea level). We carried out the following analyses: 1) careful sediment dating (210Pb dating combined with 137Cs and Ra measurements, 14C dating) and age-depth model development, 2) subsampling at 2mm resolution (and 1 cm further downcore, prior to AD 1900), 3) stomatocyst analyses (using an SEM) for each sample. The downcore variability of stomatocysts was interpreted in two ways; the Transfer Function was applied, and a calibration-in-time approach was tested. The Transfer Function provided a valuable reconstruction of winter-spring temperatures back to ca. 1000 BC. Results show good agreement between reconstructed and measured temperature variability since AD 1920, which is an independent line of evidence showing the reliability of the chosen proxy and calibration model. The alternative calibration-in-time approach did not yield conclusive results since no robust correlation was found between the raw stomatocyst data and measured temperatures. In addition, two other lakes were targeted for detailed cyst analyses. However, the cyst compositions of L. Chepical (see below) and L. La Mula presented problems due to a lack of analogy to the modern training set. Therefore we could not develop a second cold season temperature reconstruction.

The most important conclusion from this work is that recent cold season warming in this region (since AD 1980) is not unique in a 100 or 1000 year context. Although a clear warming was observed after AD 1980 which continues to the present day, this warming was not unprecedented in the context of even the last 100 years. This is in direct contrast to records from the Northern hemisphere. This highlights the importance of spatially and seasonally explicit reconstructions.

In addition, we produced a highly detailed, quantitative summer temperature reconstruction for Laguna Chepical, based on VIS-RS scanning techniques. Laguna Chepical is a glacial lake in the central Chilean Andes (32 °S), situated at 3050 m a.s.l. Scanning reflectance spectroscopy in the visible light range provided the spectral index R570/R630, which reflects the clay mineral content in lake sediments. For the calibration period (AD 1901–2006), the R570/R630 data were regressed against monthly meteorological reanalysis data, showing that this proxy was strongly and significantly correlated with mean summer (NDJF) temperatures (R3yr = −0.63 padj = 0.01). This calibration model was used to make a quantitative temperature reconstruction back to 1000 BC. Results showed that the warmest period of the past 3000 years occurred between AD 1950-1970. The same methodology was applied to sediments from L. Escondida. However, here the VIS-RS approach did not work, since the sediment core reflectance values did not display a clear (significant) summer temperature signal.

An additional project aim was to reconstruct past changes in seasonality. For Laguna Escondida, a cold season temperature reconstruction was produced back to AD 160 but the summer temperature reconstruction was not feasible. The opposite was the case for L. Chepical, as describer previously. We therefore compared the winter temperature reconstruction from L. Escondida to the summer temperature reconstruction for L. Chepical. With these data it was possible to make a tentative reconstruction of past changes in seasonality. This comparison showed that winter and summer temperature trends in the Andes differed substantially, indicating the necessity of seasonally explicit reconstructions. Seasonal contrasts changed substantially through time. Low seasonality points to a strong influence of Westerlies in the study area, whereas high seasonality can best be explained by a more southern position of the Westerly wind belt. However, due to chronological uncertainties when comparing two records from different lakes, and the substantial distance between the two lakes, we cannot draw any final conclusions on changes in seasonality without risking strong over-interpretation of the data.

This project has contributed substantially by providing 1) technological innovations (a: confirming chrysophyte stomatocysts as a reliable, quantitative proxy, b: development of training set using novel field and statistical techniques), 2) scientific knowledge (providing quantitative, detailed, seasonally explicit temperature reconstructions for a region where so far very few of these existed), and 3) training of students. Within the context of this project, one B.Sc. and two M.Sc. projects were carried out and completed successfully. The two M.Sc. students gained valuable experience in the field, laboratory and in data analysis and reporting. Results were disseminated at international conferences, workshops and during seminars. The results were so far published in three reports and two international journals.

relevant contact details: dr. Rixt de Jong, alt. Prof. Dr. M. Grosjean,