Speaker: Christina Karamperidou, University of Hawaii, Mānoa

Title: "Multi-resolution and hierarchical approaches to ENSO diversity in past, present and future climates"

Abstract: Climate model skill in simulating ENSO diversity is not only critical for accurate simulation of global ENSO impacts but may also be used as a tool to constrain future climate projections. In models that exhibit strong ENSO diversity, the projected tropical Pacific sea surface temperature warming pattern is closely linked to the projected ENSO response, highlighting that ENSO simulation biases may lead to potentially biased projections in long-term SST and precipitation trends, with great significance for regional climate adaptation strategies (Karamperidou et al. 2017). Here, I will first present a hierarchical model study to explore the factors affecting the existence of ENSO flavors or regimes. The hierarchy of models includes: a) a nonlinear version of the recharge-discharge model (Takahashi, Karamperidou & DeWitte, 2018); b) an intermediate complexity ENSO model (modified Zebiak-Cane by Xie & Jin, 2018); and c) a comprehensive earth system model (NCAR’s CESM). I will show changes in the distribution of ENSO peaks and the dynamics of the two ENSO flavors leading to their antithetical responses to Holocene forcing (Karamperidou et al. 2015). However, ground-truthing these model results using paleoclimate proxies is challenging: First, depending on the location, one or the other ENSO flavor may dominate the regional signals, and antithetical ENSO-flavor changes may result in compounded or muted signals. Second, regional climate response and especially changes in extreme event frequencies may deviate from large-scale signals due to the impacts of mesoscale circulation processes and topography. Therefore, a dynamical and statistical downscaling approach is necessary in order to assess the regional impacts of ENSO flavors and how changes in their relative frequency may be reflected in paleoclimate records. Focusing on the tropical Andes, I will combine dynamical downscaling of ENSO-induced precipitation using a mesoscale weather forecasting model (WRF) with extreme value analysis to show that intense precipitation in this region is ENSO-type specific and an altered relative frequency of ENSO flavors can help explain local paleoclimate records (Kiefer & Karamperidou, 2019). The studies I will present highlight the usefulness of a multi-resolution and hierarchical approach for providing insights into past and future ENSO behavior and the complexity of its impacts across the Pacific, with applications in the fields of paleoclimate model-proxy synthesis, present and future regional risk assessment, and the development of emergent constraints for future climate projections.

Biography: Dr. Karamperidou is an Assistant Professor at the Department of Atmospheric Sciences, U. Hawaii. She received her Ph.D. in 2012 from Columbia University, where she also holds an adjunct associate researcher position. Her research interests include ENSO dynamics and predictability, ENSO in past climated, the response of mid-latitude atmospheric circulation to climate change and variability, sea level variability and its effects on coastal aquifers and ecosystems, dynamical systems theory, hydro-climate modeling, and hydroInformatics and model optimization.