Exploring last millennium tropical Pacific hydrology: Insights from climate models, proxy data, and proxy system modeling

Climatic phenomena originating in the tropical Pacific significantly influence global hydroclimate patterns. Understanding of tropical pacific hydroclimate is difficult given the short lengths of instrumental records, which in turn hampers future climate predictions. Paleoclimate records can extend the instrumental record and further our understanding of hydroclimate in the tropical Pacific. Additionally, climate model simulations provide a means of analyzing climate dynamics and teleconnections at a variety of spatial and temporal resolutions. Using these tools, I explore the last millennium hydroclimate in the tropical Pacific at a variety of spatial scales. I present a new regional study of the El Niño Southern Oscillation (ENSO) and zonal sea surface temperatures (dSST) to improve understanding of how tropical Pacific phenomena relate to one another. General circulation model (GCM) simulations participating in the phase three Paleoclimate Modeling Intercomparison Project show diversity in the strength and direction of the ENSO-dSST relationship. Models that best simulate tropical Pacific climate tend to have a negative ENSO-dSST correlation, suggesting there is an increased likelihood of enhanced ENSO during periods of reduced tropical Pacific dSST. Additionally, all model simulations suggesting the relationship is non-stationary as the ENSO-dSST relationship changes sign with time. I developed new paleoclimate reconstructions from Lake 30 on Kiritimati (2oN, 157oW) that allow for new interpretations of central Pacific hydroclimate change over the last millennium. Total lipid d2H reconstructions from lake sediments produce the first quantitative reconstruction of salinity for Kiritimati. Paired with sediment mineralogy, carbonate isotope data, and new radiocarbon dates, results reveal an arid period from 950-1370 CE, a hiatus due to the complete evaporation of the lake from 1370-1970 CE, and a wetter period in recent decades. Comparison with sediment records from nearby Washington Island reveal consistent aridity in the Central Pacific until modern decades, when both islands become wetter. Consistent aridity in the Central Pacific is most likely caused by decreased ENSO variability or changes in the Pacific Walker Circulation. Patterns, as observed here, are unlikely to be caused by ITCZ shifts, as previously proposed. To gain further insights into tropical hydrology, I explore local controls on carbonate d18O values. Carbonate d18O reconstructions from Lake 30 sediments show substantially different multi-decadal patterns from salinity reconstructions, despite both being controlled by moisture balance changes. Carbonate d18O values generated from proxy system modeling of Lake 30 reveals the importance of evaporation and precipitation on carbonate d18O values. Lake salinity is also revealed as an important environmental control on lake evaporation rates and thus carbonate d18O values. Model difficulties simulating changes in lake salinity likely account for uncertainness during more arid periods. The importance of lake salinity on carbonate d18O values likely explains the differences between last millennium reconstructions of Lake 30 P-E and carbonate d18O values.LimitedAuthor requested closed access (OA after 2yrs) in Vireo ETD syste