by Paul A. Baker

1. The importance of the tropics and tropical climate The tropics (30°N to 30°S) account for more than one half of Earth’s land area and more than twothirds of Earth’s human inhabitants. It is estimated that 80% of all plant and animal species are found in tropical rainforests—the Amazon basin alone contains more than 30% of the area of these rainforests. Primary productivity and carbon storage by tropical rainforests comprise important fractions of the global totals, hence tropical forest biomes play crucial roles in the global cycles of carbon dioxide and methane. For these reasons, it is imperative that we completely understand the workings of tropical ecosystems and, we argue, the most crucial need is to understand tropical climate change. For example, different general circulation models (GCMs) predict radically different outcomes for tropical climate under conditions of future CO2 increases, alternatively producing wetter or drier, even desert, conditions in the Amazon basin. Which scenario is correct? Will there be changes in the frequency and amplitude of ENSO variations during future conditions of enhanced CO2? How will the tropical monsoons respond to these conditions? What will be the extratropical effects of changing monsoonal or ENSO variability? Is the oceanic thermohaline circulation sensitive to tropical climate variability? While we will never be able to perfectly forecast future climate, one of the best methods for determining what the future holds for tropical climate is to determine how climate has changed in the past, during times of different global climatic mean states. This is a topic of important societal relevance and it is the main concern of the present proposal. There are three major centers of deep atmospheric convection on Earth, tropical South America, tropical Africa, and the "maritime" continent of the western Pacific (reviewed by Hastenrath, 1991). These regions have dominant roles in energizing global atmospheric circulation through radiative and latent heating and the input of water vapor. These regions also introduce important perturbations into an otherwise more zonally-uniform circulation. Climatic variations in the tropics on interannual and interdecadal time scales affect higher latitude climate as well (e.g. Wallace et al., 1990, Rajagapolan et al., 1998) and, in turn, are influenced by higher-latitude climate (e.g. Xie and Tanimoto, 1998). Surprisingly little is known about either the character or the causes of longer (e.g. millennial) time-scale variations in tropical climate and their possible global teleconnections. We propose to study the nature and origin of past climate variation in tropical South America on timescales from multi-decadal to orbital for a period covering the last 0.5 million years. 2. Lake Titicaca as a unique sedimentary archive Lake Titicaca provides a unique opportunity to study a continuous record of past climate in tropical South America for several reasons: (1) Lake Titicaca is the only large and deep freshwater lake in South America, hence in deeper portions of the lake, sediment accumulated continuously and rapidly (~ 1m 103 yr-1) for at least a few hundred thousand years. By contrast, nearly all of the previously studied sedimentary records in the Amazon basin are discontinuous or very much lower resolution (Ledru et al., 1998). The lake is beyond the glacial limit and occupies a basin of late Pliocene or Pleistocene age. (2) Despite its high altitude (3810 masl) and moderately high latitude (about 16° to 17.50°S, 68.5° to 70°W), Lake Titicaca is a reliable recorder of the climate of a large portion of tropical South America. The annual cycle and interannual anomalies of precipitation around Lake Titicaca are correlated with those in western and central Amazonia south of the equator (most of Amazonia). Changes in lake level are well correlated with precipitation amounts (and temperature) in the Amazon basin, and tropical sea-surface temperatures (SST) in the adjacent equatorial Atlantic. (3) Modern Lake Titicaca is a nearly closed basin, thus lake level, chemical composition, and biota are particularly sensitive to changes in the amount of precipitation and the precipitation/evaporation ratio. The important climatic elements can be clearly deduced from the proxy records. (4) There are other types of climatic records in this northern Altiplano region of Bolivia and Peru, including ice cores from the Quelccaya ice cap (Thompson et al., 1986) and Volcan Sajama (Thompson et al., 1998). Comparison of the lacustrine records of paleoclimate with these ice-core records bolsters our confidence in the interpretation of both. The lacustrine records extend much farther back in time. (5) Lake Titicaca is perfectly situated to very directly record the history of glaciation in the tropical Andes. There have been few investigations of timing and amplitude of middle- to late-Pleistocene glaciation prior to the last glacial maximum (LGM)—because of erosion, evidence of these older glacial stages can only be preserved in downstream sites beyond the maximum glacial limit. (6) Lake Titicaca is also well situated to receive volcanic ash from the adjacent active arc—there are several active volcanic centers within 200 km from the lake (including the well known Sabancaya and Huaynaputina). The lake will provide an important record of volcanism and volcanic ashes will be essential for dating older sediments in our cores. (7) The lacustrine basin straddles the morpho-tectonic boundary between the eastern and western Andean cordillera. Drilling in the lake offers the possibility of dating and characterizing the nature of faulting and basement rocks of this important transition that has already been imaged in our seismic-reflection profiles.

• The major focus of our proposed research is on tropical paleoclimatic (including glacial) reconstruction. This is the goal that we feel is most societally-relevant and most scientifically compelling, for which Lake Titicaca is optimally situated, and for which drilling is the ideal (and only), capable method. Secondary, but important, goals include recovery of a record of regional Andean volcanic activity and elucidating the tectonic origin of the lake basin. We present a list of questions (hypotheses) to be addressed by the drilling of Lake Titicaca. *What is the nature of climate change in tropical South America during the past 0.5 Myr? We know from seismic records that there have been large changes in effective moisture and lake level—when were these? Were they caused by variations in the South American summer monsoon? Were they forced by the known variations of insolation at orbital time scales (especially the precessional variations at 20 kyr time scales)? What is the phasing of Amazon moisture variability with respect to the global methane signature as observed in ice cores (e.g. Chappelaz et al.,
• 1993)? *Are there Pleistocene millennial-scale changes in precipitation and temperature such as we have already observed in the Holocene and late glacial record of the lake? What is the nature of their phasing with respect to Heinrich events and Dansgaard-Oeschger cycles of the high-latitude North Atlantic region? How does the amplitude of these events in the tropics differ between glacial and interglacial stages? *What are the linkages between tropical climate change and global change? How do atmosphericallyforced wet-dry cycles in tropical South America correlate with the global and regional patterns of seasurface temperature variability on a variety of resolvable time scales, but time scales that are much longer than modern instrumental records (answers to this question may need to await high resolution SST reconstructions on marine cores throughout the Atlantic, comparable to the coverage by CLIMAP (1981), but at many different time slices and with reliable temperatures).
• Today we describe regional patterns of SST variability or atmospheric variability using numerous indices (e.g.ENSO, NAO, PADO, PDO, AO, the tropical dipole, etc.) that serve to capture much of the variability of climate for the last few decades. Do the same “centers of action” or geographic patterns of variability hold on longer time scales? Do the centers of action or patterns of variability change with different global mean states (e.g. glacial-interglacial)? *What is the timing of tropical glaciation? Our piston cores from Lake Titicaca contain a clear record of tropical glacial advance and retreat for the last glacial maximum (LGM). We can think of no other sites worldwide that are as likely as Lake Titicaca to contain a continuous and easily decipherable record of tropical glaciers prior to the LGM. What was the timing of previous glacial advances and retreats? Were these always in phase with the high-latitude “global” glacial stages? Or, as seems more likely, were they forced by major climate change (increased wetness and perhaps lower temperatures) at precessional time scales? *To what extent was the climate of tropical South America affected by changing high-latitude boundary conditions (e.g. glaciation) and global surface temperature changes? Is there any evidence for a role of the natural variability of CO2 in forcing tropical climate, perhaps through a vegetation feedback (e.g. Kutzbach et al.,
• 1996). *What is the record of volcanic activity in the late Quaternary? Are there detectable changes in the frequency of eruptions from different volcanic centers? *What is the age and nature of seismically-identified basement underlying the late Quaternary sediments at our drill sites? We expect that basement rocks will help elucidate the tectonic origin and timing of the formation of the Lake Titicaca basin. *What is the heat flow at these sites and is there any evidence for deep fluid flow?

Andes, Bolivia, Climate Change, GLAD, GLAD800, Global Environment, ICDP-2000/12, Lake Drilling, Peru, South America, TITICACA

Latitude: -16.12053, Longitude: -69.12053