Overview

Project Highlights:

  • Leading edge developments of the next generation of weather prediction tools using High Performance Computing (HPC).
  • Simulations of atmospheric processes covering an immense span of scales, from a fraction of a millimetre, from growth of cloud droplets, to planetary scale.

According to the World Metrological Organisation weather and climate-related disasters have caused $2.4 trillion in economic losses and nearly two million deaths globally since 1971 as a result of hazards such as droughts, extreme temperatures, floods, tropical cyclones, and related health epidemics. To date, improvements in numerical modelling and observation systems have significantly contributed to improving the accuracy of atmospheric forecasts.

Existing operational Numerical Weather Prediction models are, by design, not equipped to resolve convective motions which impairs the fidelity of forecasts. A new generation of numerical weather prediction tools is needed to resolve and describe a vast range of scale interactions, from organised cloud systems with deep moist convection in the atmosphere, to small scale orographic and land-use diversity, and eddies in the oceans. The fidelity of everyday forecasts and severe weather warnings in a changing climate can be improved by more accurate modelling of all thesefeatures.

This project will contribute to the development of the next generation of numerical weather prediction tools by exploiting novel unstructured mesh methodologies developed at Loughborough University under NERC funded project NERC/G004358. These methodologies have already provided important components for a recent flexible non-hydrostatic modelling strategy adopted by the European Centre for Medium Range Weather Forecasts (ECMWF) and for a research project at the US National Center for Atmospheric Research (NCAR), via DoE DE-SC0006748 “Multiscale Simulation of Moist Global Atmospheric Flows”. The proposed work, in collaboration with both institutions, will build on these developments. The project will also benefit from the High Performance Computing (HPC) environment being developed under the ESCAPE Horizon2020 project [1].

Computing flows on a sphere (Earth).

Methodology

The research will address challenging applications for which unstructured meshes have the potential to overcome limitations of regular grids, traditionally used in simulations of atmospheric flows. Furthermore, the project will contribute to advancing a new class of non-hydrostatic models for geophysical flows; opening avenues for highly accurate predictions in a wide range of environmental applications beyond atmospheric flows, e.g. floods or dynamic evolution of bathymetry in rivers.    

The project will exploit and enhance bespoke in-house mesh generators and Nonoscillatory-Forward-in-Time nonhydrostatic flow solvers operating on unstructured meshes [2]. These tools will provide highly accurate global and limited area models for simulations of a range of challenging stratified atmospheric flows. The research will concentrate on advanced simulations and numerical developments, together with rigorous analysis and validation of results.  Computations and developments will take advantage of HPC architectures for large data sets.

Fundamentally important questions can be researched using these novel explicitly multi-scale tools; such as how local phenomena (e.g. severe convective events over the US) lead to “forecast busts”, affecting the weather five days later over Europe. It would also be possible to investigate the dynamics of the stratocumulus top which is one the most challenging problems in small-scale atmospheric dynamics, and may provide key information to the way such clouds respond to climate change.

Training and Skills

CENTA students are required to complete 45 days training throughout their PhD including a 10 day placement. In the first year, students will be trained as a single cohort on environmental science, research methods and core skills. Throughout the PhD, training will progress from core skills sets to master classes specific to CENTA research themes.

The student will aid the development of a range of novel atmospheric models and advanced simulations. Therefore, the project will help the student to build a background in: Numerical Methods, Applied Mathematics, Geophysics, Fluid Dynamics, Engineering Environmental Flows and other related disciplines. If required, a full training in programming will be provided as it will be an essential tool needed in this project. Some programming background would be an advantage. The detailed project tasks can be tailored to the background and interests of the student and all the above skills will be fostered throughout the research programme by the expert LU staff and their international collaborators. Activities will potentially include placements at ECMWF and/or NCAR and attending the annual ECMWF’s course on “Advanced numerical methods for Earth system modelling”.

Timeline

Year 1: Develop alternative computational meshes for simulations of selected atmospheric problems. These may include generating meshes for the Planet Earth (global models) or meshes customised for modelling stratocumulus clouds for limited-area models.

Year 2: Customisation of non-hydrostatic unstructured mesh-based codes for the efficient implementation of challenging applications on new types of computational meshes.

Year 3: Validation and research studies for the selected range of atmospheric applications.

Partners and collaboration (including CASE)

The LU investigators and the ECMWF and NCAR scientists have been collaborating closely for over a decade. The research team’s accomplishments include development and testing of a novel all-scale family of models capable of very high resolution simulation of the Earth’s atmospheric general circulation, applying either structured grids [3] or unstructured meshes [4,5,6]. The project will benefit from ECMWF and NCAR scientific guidance and computer time on the HPC.

Further Details

For information about this project, please contact Dr Joanna Szmelter (j.szmelter@lboro.ac.uk).  For enquiries about the application process, please contact Lauren Curtis, Wolfson School of Mechanical, Electrical and Manufacturing Engineering, Loughborough University (Ws.Phdadmin@lboro.ac.uk).

Please quote CENTA when completing the application form: http://www.lboro.ac.uk/study/apply/research/.