Integrating climate model simulations and paleo data to constrain future drought risks over California

Susan Zimmerman | 17-ERD-052


Water resource managers rely on climate model predictions of future precipitation patterns to prepare for droughts and floods. However, model simulations of California's future hydroclimate have very large uncertainties, making practical decisions difficult.

We determined that extratropical climate responses to external forcings can have significant influence on California's hydroclimate. This highlights the importance of reliable representation of the linkages between forcings and responses to accurately predict future hydroclimatic changes in California. Our improved understanding of the way hemispherically asymmetric forces can impact the changes in precipitation and aridity over California, as seen in CMIP5 models, CanESM2 simulations, and observations, improves predictions of 21 st century changes in Californian rainfall.

We analyzed model simulations forced by estimated historical changes in both anthropogenic and natural factors and compared them with observations. We identified model-predicted "fingerprints" of hydroclimate change in observed temperature, rainfall, and aridity data. Over the period from 1950 to the present, one of the fingerprints we isolated captured pronounced interhemispheric temperature contrast associated with a shift in the latitude of the Intertropical Tropical Zone (ITCZ), resulting in precipitation deficits and aridity in California, the Sahel, and India. This ITCZ shift has complex temporal behavior that is influenced by emissions of both greenhouse gases and aerosols. We also found that a southward ITCZ shift is positively correlated with a wetter California. The covariance relationship between ITCZ shifts and California rainfall anomalies is consistent in the observations and climate simulations over the last 65 years.

To improve our understanding of patterns in pre-instrumental records of natural climate variability, we developed a robust and novel method for dating pollen preserved in lake sediments. We demonstrated that pollen dates can be as accurate as dates based on macroscopic plant material and applied our new dating method to three records of past climate change from lakes in California. This enabled an improved assessment of the teleconnections identified in model simulations as causes for natural variability in California's past.

Impact on Mission

This project supports the Lawrence Livermore National Laboratory's energy and climate security mission area and leverages the Laboratory's core competencies in earth and atmospheric science and isotopic science and technology. A new water science focus was established at Livermore during the research period, and project scientists were identified as expert members of this new initiative to align Laboratory capabilities in hydrological science, climate adaptation, and infrastructure innovation. The robust pollen-dating method developed represents a new capability for Livermore's Center for Accelerator Mass Spectrometry (CAMS) and, therefore, enhances the Laboratory's international reputation for innovation and high-quality chronological tools.

Publications, Presentations, Etc.

Cvijanovic, I., et al. 2017. "Future loss of Arctic sea-ice cover could drive a substantial decrease in California's rainfall." Nature Communications , v. 8 (2017). LLNL-JRNL-680635.