English (United Kingdom)

https://curated-unify.zendy.io/wp-json/zendy-region/v1/featured_content/oa?rat=en

https://curated-unify.zendy.io/wp-json/zendy-region/v1/highlighted_journal/

Global: Zendy Tools annual

Presents the keyphrase highlighting as premium feature

Keyphrase Highlighting

Presents the summarisation as premium feature

Summarisation

Insights

Presents the pdf analysis as premium feature

PDF Analysis

Presents the zaia usage as premium feature

ZAIA

Global: Zendy Tools monthly

Global: Zendy Plus monthly

Presents the access of premium content as premium feature

Premium Content

Global: Zendy Plus annual

Global: Zendy Free Plan

Abstract. Models of atmospheric phenomena provide insight intoclimate, air quality, and meteorology and provide a mechanism forunderstanding the effect of future emissions scenarios. To accuratelyrepresent atmospheric phenomena, these models consume vast quantities ofcomputational resources. Machine learning (ML) techniques such as neuralnetworks have the potential to emulate computationally intensive components of thesemodels to reduce their computational burden. However, such ML surrogatemodels may lead to nonphysical predictions that are difficult to uncover.Here we present a neural network architecture that enforces conservationlaws to numerical precision. Instead of simply predicting properties ofinterest, a physically interpretable hidden layer within the networkpredicts fluxes between properties which are subsequently related to theproperties of interest. This approach is readily generalizable to physicalprocesses where flux continuity is an essential governing equation. As anexample application, we demonstrate our approach on a neural networksurrogate model of photochemistry, trained to emulate a reference model thatsimulates formation and reaction of ozone. We design a physics-constrainedneural network surrogate model of photochemistry using this approach andfind that it conserves atoms as they flow between molecules whileoutperforming two other neural network architectures in terms of accuracy,physical consistency, and non-negativity of concentrations.

Conservation laws in a neural network architecture: enforcing the atom balance of a Julia-based photochemical model (v0.2.0)