Silicone v1.0.0: an open-source Python package for inferring missing emissions data for climate change research

Abstract

Integrated assessment models (IAMs) project future anthropogenic emissions which can be used as input for climate models. However, the full list of climate-relevant emissions is lengthy and most IAMs do not model all of them. Here we present Silicone, an open-source Python package which infers anthropogenic emissions of unmodelled species based on other reported emissions projections. For example, it can infer nitrous oxide emissions in one scenario based on carbon dioxide emissions from that scenario plus the relationship between nitrous oxide and carbon dioxide emissions found in other scenarios. Infilling broadens the range of IAMs available for exploring projections of future climate change, and hence Silicone forms part of the open-source pipeline for assessments of the climate implications of IAM scenarios, led by the Integrated Assessment Modelling Consortium (IAMC). This paper presents a variety of infilling options and outlines their suitability for different cases. We recommend certain infilling techniques as good defaults but emphasise that considering the specifics of the model being infilled will produce better results. We demonstrate the package's utility with three examples: infilling all required gases for a pathway with data for only one emission species, splitting up a Kyoto emissions total into separate gases, and complementing a set of idealised emissions curves to provide a complete, consistent emissions portfolio. The code and notebooks explaining details of the package and how to use it are available on GitHub (https://github.com/GranthamImperial/silicone, last access: 2 November 2020). The repository with this paper's examples and uses of the code to complement existing research is available at https://github.com/GranthamImperial/silicone_examples (last access: 2 November 2020).