CHAP: 2017 Accelerated Scientific Discovery

The Accelerated Scientific Discovery (ASD) initiative provides dedicated, large-scale computational resources to a small number of projects for a very short time period, usually two months or so following acceptance of a new HPC system. These projects are selected to help put the new system through its paces and pursue scientific objectives that would not be possible through normal allocation opportunities. In most cases, three to five NSF-supported, university-led projects from across the geosciences or the supporting computational sciences are chosen.

For the Cheyenne HPC system, the second system to be installed at the NCAR-Wyoming Supercomputing Center (NWSC), CISL issued calls for university and NCAR proposals in mid-2016 and received 11 proposals for ASD projects, ranging from 10 million core-hours to 58 million core-hours. Altogether the requests asked for 300 million core-hours, but only 250 million core-hours were available. The university proposals were reviewed by the CHAP, and the following five awards were made.

Project Metis: Seasonal forecasts with enhanced ocean and atmosphere resolution

Project lead: Benjamin Cash, George Mason University 
Cheyenne allocation: 58 million core-hours

Project Metis is a natural continuation of the highly successful collaboration between the Center for Ocean-Land-Atmosphere Studies (COLA) and the European Centre for Medium-Range Weather Forecasts (ECMWF), in support of both centers’ ongoing efforts to understand and quantify predictability in the climate system from daily to interannual time scales. Metis represents a follow-on to the Athena and Minerva projects, which focused on understanding the impact of atmospheric resolution on seasonal prediction. Metis will explore the impact of increasing both atmospheric and oceanic resolution on model fidelity and prediction skill in a coupled ocean (waves, currents and sea-ice), land, and atmosphere ensemble framework. We will test very high-resolution model configurations, with grid spacing up to 9 km in the atmosphere and 25 km in the ocean, which represents a major step forward with respect to current seasonal and sub-seasonal forecasting systems.

Estimating CESM high-res near-term climate predictability

Project lead: Benjamin Kirtman, University of Miami
Cheyenne allocation: 17 million core-hours

This proposal seeks to provide a comprehensive assessment of how ocean eddies affect decadal predictability; and strategies for initializing ocean eddy-resolving prediction systems. It will test these strategies in experimental decadal prediction. In order to meet this objective, we will develop and implement a set of novel diagnostic and prognostic predictability experiments with a state-of-the-art ocean eddy-resolving earth system model.

Impacts of smoke aerosols on the transitions between closed and open cellular convections of stratocumulus clouds over southeast Atlantic

Project lead: Xiaohong Liu, University of Wyoming
Cheyenne allocation: 21.5 million core-hours

Stratocumulus clouds are extremely important since they cover 23% of the Earth’s surface, more than any other types of clouds. Since stratocumulus clouds strongly reflect solar radiation while only exert a small effect on outgoing longwave radiation, they can introduce strong negative cloud radiative forcings (cooling effect). Small changes in cloud coverage and thickness in stratocumulus, such as transition between closed and open cellular cloud patterns as observed by satellites, can generate a radiative effect comparable to the effect of greenhouse gases [Wood, 2012]. The closed-cell state is characterized by weak updrafts in the broad cloudy cell center and stronger, narrower downdrafts around the cell edges. In contrast, open-cell state is characterized by narrow, strong, cloudy updrafts surrounding broad, weak, clear downdrafts in the center of cell [Feingold et al., 2015]. Apparently, such sharp contrast in the cloud fields between two patterns causes a large climatic effect, since closed cellular clouds are much more reflective than open ones. Despite the importance, the transition between closed and open cells cannot be modeled by global climate models (GCMs), but only properly captured by large-eddy simulations (LES). This is because stratocumulus is heavily modulated by turbulence-induced entrainment, which can only be explicitly and accurately modeled by LES [Wood, 2012].

Previous LES modeling studies suggest that the transition between two cloud patterns can be affected by precipitation oscillations [Feingold et al., 2010], loss and restore of turbulence kinetic energy [Feingold et al., 2015], and spatial distribution of precipitations [Yamaguchi and Feingold, 2015]. Previous studies also demonstrate that aerosol particles can significantly affect the properties of stratocumulus clouds (e.g., cloud liquid water, cloud fraction, and entrainment at cloud tops) through aerosol microphysical effect (e.g., acting as cloud condensation nuclei) [Ackerman et al., 2004; Bretherton et al., 2007]. However, the influence of aerosols on the transition between two cloud patterns remains highly uncertain. As a consequence, radiative forcing of aerosols due to aerosol-cloud interactions remains one of the largest uncertainties in the future climate projection by GCMs [IPCC, 2013].

The overall goal of this study is to examine to what extent biomass burning (BB) aerosols and heterogeneity of BB aerosols affect the frequency of transition between closed and open cellular clouds, macro- and microphysical properties of stratocumulus clouds (e.g., cloud water path, cloud top height, etc.), and associated cloud radiative forcings.

Computational study of wind turbine performance and loading response to turbulent atmospheric inflow conditions area of interest: Fluid dynamics and turbulence

Project lead: Dimitri Mavriplis, University of Wyoming
Cheyenne allocation: 15 million core-hours

Predicting wind farm performance represents a complex problem that spans spatial and temporal scales over eight orders of magnitude from the continental scales that govern wind patterns to the thin boundary layers over the wind turbine blades. Improved prediction of wind farm productivity requires good resolution of flow structures and reliable modeling of turbulent eddies in this entire length-scale range. Structural mechanics further compound this problem by introducing yet another set of temporal scales such as those due to significant elastic vibrations of the rotor blade and masts, which can be one to two orders of magnitude smaller than the wind turbine rotor period. At UW, we have developed a high-fidelity, multi-scale modeling methodology that can accurately predict the performance of wind turbines by reliably modeling the entire range of these spatial and temporal scales. The developed software framework uses the large-eddy simulation (LES) approach for prediction of the turbulent flow fields. A multi-mesh framework is used to provide accurate and efficient prediction capabilities for vortex-dominated wind turbine flow fields. In addition to performing simulations for validation and predictive purposes, we are continuing the development of this capability through the addition of higher-order discretizations, and an additional level of coupling by using an intermediate LES capability for atmospheric boundary layer (ABL) flow in complex terrain at scales larger than those of the high-fidelity wind-turbine/farm simulations but smaller than the WRF simulations. The request covers time allocation for a large research group that includes four faculty and ten postdocs and graduate students.

Turbulence and magnetic reconnection in the heliosphere

Project lead: Michael Shay, University of Delaware
Cheyenne allocation: 18.3 million core-hours

Plasma turbulence and magnetic reconnection play a fundamental role in space weather. Magnetic reconnection is an explosive release of magnetic energy that is believed to help create coronal mass ejections (CMEs) and solar flares, and couples the solar wind to the Earth's magnetosphere to drive space weather in near-Earth space. Plasma turbulence, on the other hand, plays an important role in heating the solar corona and the solar wind, the drivers of space weather. A critical role that these two processes perform is to convert flow and magnetic field energy into plasma heating and energetic particles. However, the Sun-Earth system is not in thermodynamic equilibrium, so simple fluid closure models are inadequate; a fully kinetic treatment is required. In this study we perform the first comprehensive and self-consistent three-dimensional study of how plasma turbulence and reconnection dissipate energy into plasma heating. Besides examining reconnection and turbulence individually, we also explore their symbiotic relationship by cross-comparing our reconnection predictions with the dissipation observed in turbulence. If successful, this project will significantly enhance our understanding of processes that form a most basic part of space weather phenomena.