Quantitative Reconstruction of Past Drought Patterns in Western North America Using Lakes, Stable Isotopes, and Modeling - Collaborators include Byron Steinman, Mike Mann, Joe Stoner, Joe Ortiz, Pratigya Polissar, Dan Nelson, Nathan Stansell, Darren Larsen, Rob Hatfield, David Pompeani, Aubrey Hillman, Matt Finkenbinder, Lesleigh Anderson, and Bruce Finney among others

Drought in the American west can have serious economic consequences through major impacts on water resources, agriculture, hydropower and fisheries. Documenting the timing, magnitude, duration and geographic pattern of past wet/dry cycles is an important first step toward understanding the causes of drought. The next, and perhaps most critical, step requires the application of modeling studies to (1) quantitatively interpret existing proxy data, (2) identify and rank the myriad factors that influence the frequency and intensity of aridity patterns such as synoptic teleconnections involving both ENSO/tropical Pacific climate dynamics and the AMOC, and most importantly (3) reduce uncertainty in probabilistic projections of future drought pattern responses to anthropogenic forcing. To this end, great progress has been made using tree-ring studies, which provide quantitative information about drought variability in western North America. However, because tree-rings are limited by the age of the chronologies (few extend beyond 500 years BP) and the statistical methods used to detrend them, inter-proxy comparisons are warranted to develop robust reconstructions of century scale trends. Lake sediment oxygen isotope records are one such comparative proxy. While suffering from their own set of limitations (e.g., relatively imprecise age control), lake sediment records have the advantage of spanning longer time periods (generally the full Holocene or more), and in the right hydrologic settings, can be quantitatively interpreted using mass balance modeling methods to produce probabilistic hydroclimate reconstructions.

Ongoing work seeks to produce multiple, quantitative paleoprecipitation reconstructions spanning the last 2000 years using consistent methods on similar lakes in the western cordillera of North America. We are currently developing lacustrine stable isotope records in Oregon, Montana, Utah, New Meixco, and British Columbia in lakes with geological and limnological characteristics similar to those of lakes previously studied in Washington. Maintaining consistency between study lakes in terms of size, hydrologic characteristics and sediment composition is a priority of the proposed research, as it will enable the application of the same methodologies at all sites so that the resulting datasets are directly comparable. Variations in the open-basin isotope records are being combined with the closed-basin records to produce a reconstruction of drought variability that is not influenced by changes in the isotopic composition of precipitation. We are using hydrologic and isotope mass balance models to quantitatively interpret the lake sediment oxygen isotope and lake level records and thereby develop regional precipitation reconstructions. These reconstructions are used to investigate hypotheses regarding contrasting, north-south aridity patterns in the Pacific NW and the desert SW during the Medieval Climate Anomaly and the Little Ice Age and, furthermore, to assist in ongoing experiments designed to constrain key climate parameters through perturbed physics ensemble runs of a coupled ocean-atmosphere model. These parameters govern important potential changes in ENSO, the AMOC, and their associated impacts on extratropical precipitation and drought patterns in response to future anthropogenic climate forcing.

Resulting publications include

Drought Variability in the Pacific Northwest from a 6,000-yr Lake Sediment Record. We present a 6,000-yr record of changing water balance in the Pacific Northwest inferred from measurements of carbonate _18O and grayscale on a sediment core collected from Castor Lake, Washington. This subdecadally resolved drought record tracks the 1,500-yr tree-ring-based Palmer Drought Severity Index reconstructions of Cook et al. [Cook ER, Woodhouse CA, Eakin CM, Meko DM, Stahle DW (2004) Science 306:1015-1018] in the Pacific Northwest and extends our knowledge back to 6,000 yr BP. The results demonstrate that low-frequency drought/pluvial cycles, with occasional long-duration, multidecadal events, are a persistent feature of regional climate. Furthermore, the average duration of multi- decadal wet/dry cycles has increased since the middle Holocene, which has acted to increase the amplitude and impact of these events. This is especially apparent during the last 1,000 yr. We suggest these transitions were driven by changes in the tropical and extratropical Pacific and are related to apparent intensification of the El Nino Southern Oscillation over this interval and its related effects on the Pacific Decadal Oscillation. The Castor Lake record also corroborates the notion that the 20th century, prior to recent aridity, was a relatively wet period compared to the last 6,000 yr. Our findings suggest that the hydroclimate response in the Pacific Northwest to future warming will be intimately tied to the impact of warming on the El Nino Southern Oscillation.

1,500 Year Quantitative Reconstruction of Winter Precipitation in the Pacific Northwest. Multiple paleoclimate proxies are required for robust assessment of past hydroclimatic conditions. Currently, estimates of drought variability over the past several thousand years are based largely on tree-ring records. We produced a 1,500-y record of winter precipitation in the Pacific Northwest using a physical model-based analysis of lake sediment oxygen isotope data. Our results indicate that during the Medieval Climate Anomaly (MCA) (900-1300 AD) the Pacific Northwest experienced exceptional wetness in winter and that during the Little Ice Age (LIA) (1450-1850 AD) conditions were drier, contrasting with hydroclimatic anomalies in the desert Southwest and consistent with climate dynamics related to the El Nino Southern Oscillation (ENSO) and the Pacific Decadal Oscillation (PDO). These findings are somewhat discordant with drought records from tree rings, suggesting that differences in seasonal sensitivity between the two proxies allow a more compete under- standing of the climate system and likely explain disparities in inferred climate trends over centennial timescales.

Ocean-Atmosphere Forcing of Centennial Hydroclimate Variability in the Pacific Northwest. Reconstructing centennial timescale hydroclimate variability during the late Holocene is critically important for understanding large-scale patterns of drought and their relationship with climate dynamics. We present sediment oxygen isotope records spanning the last two millennia from 10 lakes, as well as climate model simulations, indicating that the Little Ice Age was dry relative to the Medieval Climate Anomaly in much of the Pacific Northwest of North America. This pattern is consistent with observed associations between the El Nino-Southern Oscillation (ENSO), the Northern Annular Mode, and drought as well as with proxy-based reconstructions of Pacific and Atlantic ocean-atmosphere variations over the past 1000 years. The large amplitude of centennial variability indicated by the lake data suggests that regional hydroclimate is characterized by longer-term shifts in ENSO-like dynamics and that an improved understanding of the centennial timescale relationship between external forcing and drought is necessary for projecting future hydroclimatic in western North America.

Isotopic and Hydrologic Responses of Small, Closed Lakes to Climate Variability: Hydroclimate Reconstructions from Lake Sediment Oxygen Isotope Records and Mass Balance Models. Hydroclimate sensitivity simulations were conducted with a lake-catchment hydrologic and isotope mass balance model adapted to two small, closed lakes (Castor and Scanlon) located in the Pacific Northwest. Model simulations were designed to investigate the combined influences of persistent disequilibrium, reddening, and equifinality on lake water and sediment (i.e., biogenic and endogenic carbonate mineral) oxygen isotope (d18O) values and to provide a basis for quantitative, probabilistic climate reconstructions using lake sediment d18O records. Simulation results indicate that within closed-basin lakes changes in long-term (i.e., multi-decadal) precipitation amounts produce inconsistent responses in lake water and sediment d18O values that are strongly influenced by lake basin outseepage and morphometry. Simulations of variable initial conditions in which randomly generated monthly climate data (i.e., precipitation, temperature, and relative humidity) were used to force the model during the equilibration period (which precedes the application of instrumental climate data) demonstrate that Cas- tor Lake and Scanlon Lake have a somewhat limited isotopic 'memory' of ~10 years. Additional tests conducted using a Monte Carlo ensemble (in which random climate data were used to force the model) combined with d18O analyses of water samples collected from 2003 to 2011 AD, indicate that within small, closed lakes in the Pacific Northwest November-February precipitation is the strongest seasonal, climatic control on sediment oxygen-isotope values. Further, a Monte Carlo based reconstruction of 20 year average November-February precipitation amounts strongly correlates (R2 = 0.66) to instrumental values from the 20th century (with all observed values falling within modeled 95% prediction limits), indicating that probabilistic, quantitative paleoclimate interpretations of lake sediment d18O records are attainable.

Isotopic and Hydrologic Responses of Small, Closed lakes to Climate Variability: Comparison of Measured and Modeled Lake Level and Sediment Core Oxygen Isotope Records. Simulations conducted using a coupled lake-catchment isotope mass balance model forced with continuous precipitation, temperature, and relative humidity data successfully reproduce (within uncertainty limits) long-term (i.e., multidecadal) trends in reconstructed lake surface elevations and sediment core oxygen isotope (d18O) values at Castor Lake and Scanlon Lake, north-central Washington. Error inherent in sediment core dating methods and uncertainty in climate data contribute to differences in model reconstructed and measured short-term (i.e., sub-decadal) sediment (i.e., endogenic and/or biogenic carbon- ate) d18O values, suggesting that model isotopic performance over sub-decadal time periods cannot be successfully investigated without better constrained climate data and sediment core chronologies. Model reconstructions of past lake sur- face elevations are consistent with estimates obtained from aerial photography. Simulation results suggest that precipitation is the strongest control on lake isotopic and hydrologic dynamics, with secondary influence by temperature and relative humid- ity. This model validation exercise demonstrates that lake-catchment oxygen isotope mass balance models forced with instru- mental climate data can reproduce lake hydrologic and isotopic variability over multidecadal (or longer) timescales, and therefore, that such models could potentially be used for quantitative investigations of paleo-lake responses to hydroclimatic