I would like to hire three new senior members in my group at the University of California, Irvine. My main research interest is in advancing climate science by exploiting emerging atmospheric modeling algorithms of opportunity, including (1) turbulence- or cyclone-permitting variants of global cloud superparameterization that are just now becoming possible on next-gen (e.g. GPU-based) supercomputers, and (2) machine learning emulation of atmospheric physics for interpreting and accelerating complex climate system dynamics (e.g. cloud interactions with the environment). Funding for these new positions falls across 3 separate projects:
- Achieving the potential of ultraparameterization to confront issues of boundary layer turbulence in cloud feedbacks to climate change [Funding: NSF, collaborative with Chris Bretherton & Peter Blossey @ U. Washington]. Over the past five years a UCI/UW collaboration has paved the way for ultra-high-resolution (i.e., boundary layer eddy-permitting) global superparameterization (Parishani et al. 2017,18), but the potential and scientific ramifications of this have only begun to be explored. I welcome a new group member interested in understanding what relaxing traditional assumptions about the role of boundary layer turbulence means for broader climate dynamics, including the transitions and teleconnections between various forms of low cloud organization and especially the aerosol-cloud indirect effect. For this position, expertise in global atmospheric dynamics, or the modeling or observation of aerosol-cloud interaction, is desired. Synergistic interest in testing prototype algorithms and exploring next generation supercomputers is beneficial; training on high performance computing is available.
- Exploring tropical cyclone or diurnal cycle dynamics at an unfamiliar limit of global superparameterization (or machine learning emulation of same) [Funding: DOE, collaborative with many at LLNL, PNNL].Funding for this position is through my role as one of two University Partners in the DOE’s Exascale Computing Project (ECP). It is an exciting time since the DOE-ECP has just succeeded in their 3-year mission to port the embedded cloud-resolving model (CRM) calculations used in superparameterization to GPU. The reward is a newfound ability to run on the world’s best supercomputer (Summit, 200-petaflops), where previously unimaginable limits of cloud-resolving model resolution can be explored. At UC Irvine I am looking for a scientist interested in engaging with this great team of DOE researchers by helping analyze their first prototype model (i.e., the first 25-km exterior resolution superparameterized global model, with one million embedded CRMs), including its unfamiliar emergent diurnal cycle and hurricane dynamics. While the main goal is to do phenomenological and climate-change science, interest in helping with the DOE’s vision of making this prototype into an operationally-useful convection-permitting model is helpful. Testing the viability of machine learning to emulate these hard-won physical results using deep learning algorithms at scale on multiple nodes of Summit could be an attractive alternate topic under this umbrella.
- Machine learning emulation of aerosol physics and aerosol-cloud interaction [Funding: DOE, collaborative with PNNL]. The DOE has ambitious plans to augment the aerosol and cloud physics modules in its next-generation climate models — including its prototype global cloud resolving model that will run effectively on GPUs. As part of a great and growing team led by Po-Lun Ma and others at PNNL (project title: EAGLES), I am seeking someone interested in helping with this effort by exploring whether machine learning emulation of aerosol-cloud interaction calculations is a viable strategy for porting these costly calculations to GPU, and as a dynamical systems summary tool of the associated physics. Candidates with existing domain expertise on aerosol-cloud nucleation physics or aerosol cloud interaction are especially encouraged. Our group has had some recent success in emulating non-aerosol aspect of sub grid climate model physics (e.g. Rasp et al. 2018; Beucler et al., 2019) so in-house training on the practicalities of a deep learning workflow is available within the group already.
I am flexible about logistics since I am mostly hunting for talent and meaningful collaboration with fun, independent people. For candidates interested in more than one of the above topics, I am open to creative project fusions spanning more than one theme. Start dates are flexible, as is the location of work (i.e., I am open to mutually agreeable balances of on-site vs. off-site work where appropriate). The terms of appointment are typical — one year initially, renewable for up to two more years, subject to satisfactory progress; funding is in place. For applicants looking to establish new funding, it is worth knowing that it is possible at UCI for Postdoctoral scholars to lead their own proposals and that positions beyond Project Scientist exist.
Our campus is nestled in Southern California between Los Angeles and San Diego, in the heart of Orange County. It is a pleasant place to live, with a Mediterranean climate and good access to such attraction as: beaches (e.g., Crystal Cove and Laguna Beach), the San Bernardino mountains and forests, the Mojave desert and Los Angeles. The Earth System Science Department at UC Irvine is a highly interdisciplinary and modern research environment comprising ~ 25 faculty with expertise across many components of the Earth System, including atmospheric physics, land surface processes, climate dynamics, terrestrial and marine biogeochemical cycles, ice sheets, and human systems. The University of California is known for offering competitive retirement savings, health and family benefits, and has a strong institutional commitment to inclusive excellence and diversity.
If you are interested in any of these positions, please send me a brief e-mail indicating which one(s) [and why] along with your CV. I can be reached at mspritch@uci.edu.
Thanks!
Mike Pritchard
Recent Comments