Paleoclimate Records

Why use paleoclimate records?

Before I started my graduate work, when I imagined the Amazonian rainforest I envisioned an unwaveringly humid region covered with misty, low clouds, towering trees, and giant leaves. Although many parts of Amazonia are usually quite wet, the seasonal rainfall can weaken in this region. For example, southern Amazonia experienced severe droughts in 2005. 2007, and 2010. Pacific ocean surface temperatures (El Nino, or ENSO) and tropical Atlantic temperatures tend to control rainfall variations over Amazonia. During 2005 and 2010, the tropical North Atlantic ocean was unusually warm. These warm sea-surface temperatures likely anchored the tropical rainfall belt to the north, causing an unusually dry dry season across much of southern Amazonia (see figure below). Unusually dry periods like the one in 2010 lower river levels, dry out vegetation, and make the forest vulnerable to human-caused fire.

I am interested in answering questions such as: Are these droughts common? How bad can these droughts get? Our instrumental record (such as weather station or satellite observations) is quite short. It is difficult to determine how likely a drought is in a given region with a relatively short climate record. So, it is nearly impossible to use this short instrumental record to determine if the 2005 drought was a 'once-in-a-century' drought or a 'once-in-a-millennium' drought. Using climate model data and paleoclimate records, I hope to shed some light on the likelihood of severe dry periods in regions such as Amazonia.

Paleolimnology (Reconstructing past climate variability using lake sediments)

We load our backpacking bags with lightweight inflatable rafts, meter-long clear plastic tubes, and valved coring devices. After flying half way around the globe, we hike our equipment into isolated, closed-basin lakes. Anchored out in the middle of these lakes, we lower our empty plastic tubes to the lake bottoms, pound in these cylinders, and pull up cores of layered mud. We take note of our observations of the geology and hydrology of the lake catchment, carefully preserve the sediment-water interfaces of our sediments, leave temperature probes, and hike our gear and samples back to the closest road.

Drought Variability in Paleoclimate Records

Once we have collected, split, dated, and analyzed our cores, we can try to draw some conclusions about what different types of variability in our mud may tell us about past climate. First, we compare our records to instrumental (weather station, satellite) data in order to verify that our record is actually recording rainfall variability. For example, elements in one of our lake records from Peru tend to vary in-sync with Atlantic sea-surface temperatures. Observations during the satellite era suggest that Atlantic sea-surface temperatures tend to control the location of the rainfall belt over the American tropics.

After returning to our lab, we split open and photograph our cores, conduct cesium-137, lead-210, and carbon-14 dating, and begin analysis. We then shine x-rays on our sediments to determine the variation in elements down core, conduct grain size analysis, and isolate organic biomarkers (such as plant leaf waxes or bacteria cell wall bi-products) at regular depth intervals in our cores to help shed some light on past temperatures.