1 of 1

Interactive Visualization for Atmosphere Modeling�With the PAESCAL Partnership (BER)

Scientific Achievement

We used interactive visualization to help characterize atmospheric conditions for the occurrences of unrealistic, ultra-low cloud droplet number concentrations in DOE’s Energy Exascale Earth System Model (E3SM).

Significance and Impact

Ultra-low cloud droplet number concentrations are an important symptom related to large uncertainties in the simulation of aerosol’s impacts on global climate change. Our work has shed light on previously unrecognized research gaps in this area.
This interdisciplinary collaboration involving Industry partners demonstrates that holistic views provided by interactive visualization tools like ParaView can help inspire scientists and speed up their investigation of intricate problems.

A snapshot of clouds in an 2D slice of the Earth’s atmosphere within the Arctic Circle. Color shading shows the cloud amount. Magenta boxes indicate where ultra-low droplet number concentrations are found in the E3SM simulation. Snapshots like this from ParaView reveal that unrealistic results may prevail large core areas of low-altitude stratus clouds, suggesting that more research is needed to understand the life cycle of such clouds and their representation in numerical models like E3SM.

Technical Approach

New ParaView plugins were developed using Python and NumPy. This approach enables fast prototyping and development while still using ParaView’s rich analysis and visualization capabilities.
Novel approaches to visualize data, allowing mapping to a globe and applying different map projections enabled both 2D and 3D visualization for detailed insights.

PI(s)/Facility Lead(s): Berk Geveci (Kitware PI, RAPIDS); Hui Wan (PNNL, PAESCAL PI)

Collaborating Institutions: Kitware Inc., PNNL

ASCR Program: SciDAC

ASCR PM: Lali Chatterjee, Kalyan Perumalla

Greenland

Baffin Bay

LOCAL LAB POC: Berk Geveci, Kitware Inc.

TALKING POINTS:

Numerical models for climate prediction entail extensive computer programs tasked with solving highly complex mathematical and physical equations. The critical endeavor of identifying model deficiencies and enhancing their reliability demands considerable time and resources.
This interdisciplinary collaboration involving Industry partners demonstrates that holistic views provided by interactive visualization tools like ParaView can help inspire scientists and speed up their investigation of intricate problems.
In this specific example, our collaborators investigated the phenomena of unrealistic, ultra-low droplet number concentrations within DOE’s Energy Exascale Earth System Model (E3SM). Exceedingly low droplet number concentrations have long been associated with significant uncertainties in simulating aerosol’s impacts on global climate change. Our work has shed light on previously unrecognized research gaps in this area, pointing out that more efforts are needed to understand and accurately simulate the life cycle of low-altitude stratus clouds.

METADATA:

Name of the associated awarded project:

PI name(s): Hui Wan (PAESCAL PI, PNNL), Berk Geveci (Kitware PI for PAESCAL)

Project title: Physical, Accurate, and Efficient Atmosphere and Surface Coupling Across Scales

Name of the program manager: Supporting technology supported by RAPIDS through ASCR (Kalyan Perumalla, Lali Chatterjee)