2 results
Physics-informed learning of aerosol microphysics
- Paula Harder, Duncan Watson-Parris, Philip Stier, Dominik Strassel, Nicolas R. Gauger, Janis Keuper
-
- Journal:
- Environmental Data Science / Volume 1 / 2022
- Published online by Cambridge University Press:
- 28 November 2022, e20
-
- Article
-
- You have access Access
- Open access
- HTML
- Export citation
-
Aerosol particles play an important role in the climate system by absorbing and scattering radiation and influencing cloud properties. They are also one of the biggest sources of uncertainty for climate modeling. Many climate models do not include aerosols in sufficient detail due to computational constraints. To represent key processes, aerosol microphysical properties and processes have to be accounted for. This is done in the ECHAM-HAM (European Center for Medium-Range Weather Forecast-Hamburg-Hamburg) global climate aerosol model using the M7 microphysics, but high computational costs make it very expensive to run with finer resolution or for a longer time. We aim to use machine learning to emulate the microphysics model at sufficient accuracy and reduce the computational cost by being fast at inference time. The original M7 model is used to generate data of input–output pairs to train a neural network (NN) on it. We are able to learn the variables’ tendencies achieving an average $ {R}^2 $ score of 77.1%. We further explore methods to inform and constrain the NN with physical knowledge to reduce mass violation and enforce mass positivity. On a Graphics processing unit (GPU), we achieve a speed-up of up to over 64 times faster when compared to the original model.
Three - Implicit values: uncounted legacies
- Edited by Keri Facer, University of Bristol, Kate Pahl, Manchester Metropolitan University
-
- Book:
- Valuing Interdisciplinary Collaborative Research
- Published by:
- Bristol University Press
- Published online:
- 05 April 2022
- Print publication:
- 05 April 2017, pp 65-84
-
- Chapter
- Export citation
-
Summary
Introduction
University–community collaborations are often complex, fraught, emotional affairs. Participants devote a lot of time, energy and emotion to bridging differences, improvising solutions, and making things work. This can be difficult and sometimes frustrating, but can also have a transformative legacy for the participants and the wider communities they are part of. These legacies, however, are not always easy to observe, identify and authorise. As we will explore in this chapter, some of the most important legacies of community–university partnerships are intangible and refer to emotions, affects, ongoing processes and emerging potentials: for example, inspiration, confidence, friendship, as well as knowledge, ideas and networks. These legacies are at least as important as projects’ harder, more tangible and easily measurable legacies.
Our exploration of legacies started with a shared interest in the role that values play in collaborative research, and in the way in which we understand related outcomes. Exploring this through the concept of legacy was particularly relevant as it allows for a more fluid understanding, and one that can be shaped by the local project context. Thus, the theoretical starting point for this work was that making the values within collaborative projects explicit would allow for the identification and evaluation of those, ‘less tangible’, legacies. Our University of Brighton authors Harder, Burford and Hoover previously established that a values-based approach could be very successful for evaluating ‘intangible’ outcomes and achievements projects led by civil society organisations (Burford et al, 2013). They brought the approach, named WeValue, as a raw starting point to the members of two complex partnerships called Scaling Up Co-Design and the Authority Research Network (ARN), and then collectively as a consortium we co-explored, co-developed and co-generated a localisable, values-based approach for a new purpose: to identify and legitimise legacies (not only outcomes) from partnership projects (not projects from a single group or organisation).
By ‘starting from values’, we mean starting with what participants consider valuable, meaningful and worthwhile in the context of their group or partnership. An explicit values lens is first locally constructed, and then used to view, identify and evaluate legacies. The WeValue approach was previously developed to allow a formal, rigorous evaluation of ‘soft’ or ‘intangible’ achievements.