DICE-MAT

General Scope and Connection with Climate Mitigation

Introduction

Increasing global population and wealth have led to a rapid rise in material consumption, causing resource depletion and waste generation, as well as high levels of energy use and related greenhouse gas (GHG) emissions from the production and use of materials. The circular economy (CE) has been proposed as an alternative to the linear economic model and aims to reduce primary material consumption and material wastes by narrowing loops, slowing loops, and closing loops. While climate scenarios already consider energy efficiency and low carbon energy sources, the Paris Agreement goals are getting out of sight. CE strategies may be an additional lever for GHG emission reductions (related to SDG 13).

While cost-benefit Integrated Assessment Models (IAMs), such as DICE, are used to assess climate policy strategies by determining cost-optimal GHG emission reduction pathways, they do not represent the impact potential of CE strategies for climate change mitigation pathways.

We address this gap by further developing MIMOSA, a simple and transparent cost-benefit IAM covering relevant technological and socio-economic (yet no material or CE) dynamics similar to DICE, towards a global meta-model that describes the key dynamics in which CE strategies impact climate change mitigation efforts. MIMOSA has been developed and used in existing research to examine the interaction of the most significant factors, such as socio-economic developments, climate system uncertainty, damage estimates, mitigation costs and discount rates. We build on and aim at extending the existing model by introducing industry sector dynamics and integrating a novel ‘stylized’ CE modelling component. Furthermore, these dynamics will be assessed globally as well as across generic world regions. The model does not generate new results but uses data from existing scenarios calculated by other models to determine cost-benefit dynamics. To achieve this goal, we leverage the global climate and resource databases of IPCC and IRP, as well as other macro-level material/CE modelling data from existing research.

The reason for focusing on industry particularly is that material production is responsible for a large part of GHG emissions. Studying industry dynamics thus is key for understanding the impact of CE on climate mitigation overall which may bring the Paris Agreement goals back in sight. This model explores these dynamics across global regions in a first highly aggregated attempt and serves as a discussion starter for further model developments and calibrations.

Model Scope

Overall Objective and Approach

The MIMOSA model is a highly abstract, DICE-like growth model designed to analyze the effects of climate change mitigation measures on GHG emissions reductions and their socioeconomic implications. The primary objective of the MIMOSA model is to understand the temporal socio-economic-climatic dynamics resulting from emissions and emissions mitigation actions. MIMOSA does not generate new emissions or mitigation-related outcomes. Instead, it is calibrated against existing global emissions and mitigation data, enabling the exploration of interesting scenarios and dynamics related to traditional mitigation measures and, through envisioned extensions, also CE measures.

Optimization Model

MIMOSA is an optimization model that maximizes the net present value (NPV) utility of consumption by optimizing yearly emissions and carbon prices. This optimization process is achieved using a calibrated marginal abatement cost (MAC) curve, which links emissions reductions to the corresponding carbon prices.

Schematic overview of extended MIMOSA model

Existing model structure

The MIMOSA model has a global focus, providing insights into the macroeconomic dynamics and emissions reductions at a global scale. Further, extensions to incorporate regional and country-specific analyses have been developed in the existing version. The time horizon of the MIMOSA model extends until 2100, allowing for long-term projections and policy evaluations. The existing version considers the economy as a whole without distinguishing any sectors.

The existing model structure is highly generic, with macroeconomic developments driving emissions, which, in turn, cause damages to the economy. The model considers non-CE emissions mitigation interventions dependent on a variable carbon price which is derived from the MAC curve, calibrated against AR6 data. In a given year $t$, baseline emissions are reduced based on a reduction factor derived from the MAC curve: $emissions_{t}=baseline_{t}*reduction_{t}$ . Mitigation cost are calculated for the chosen carbon price and reduction factor.

Extensions for Circular Economy Dynamics

To enhance the model's analytical capabilities, efforts are being made to introduce CE dynamics in the industry sector across global regions. We are following an iterative approach and additional dynamics may be added in future iterations related to energy, materials and additional sectors.

Emissions are broken down into two sectors: industry ($ind$) and non-industry ($non-ind$). Emissions of both sectors are mitigated through sector MAC curves that are calibrated against AR6 data. Additionally, for the industry sector, we consider the impacts of CE on the carbon intensity of materials through an additional CE MAC curve that is calibrated against indications of emission reductions and associated cost from the extant literature. As such, there is no explicit material factor modeled, instead (following the high abstraction logic of cost-benefit IAMs) impacts from CE strategies on industry emissions are implicitly considered on an aggregated level. This results in two sector emission equations per region:

$emissions_{non-ind,t}=baseline_{non-ind,t}*reduction_{non-ind,t}$

$emissions_{ind,t}=baseline_{ind,t}*reduction_{CE,ind,t}*reduction_{non-CE,ind,t}$

MAC curves are linked through a common carbon price. CE MAC curve is currently under development: To build a dynamic optimization model, the cost of implementing CE measures are also required, besides emissions reduction impacts. Depending on data availability, this may be represented through scenario parameter instead.

Model Development

• Status: in progress
• Environment: Python
• Documentation: in progress (will be made publicly available once the implementation is mature)
• Source code: in progress (will be made publicly available once the implementation is mature)

Circular Economy Features

R Words coverage and implemented in the model

Being a global and highly abstract model, the CE extensions discussed for implementation are of a broader perspective as well. The current approach will not distinguish any value retention strategies explicitly (R strategies). They will be implicitly considered on an aggregated level through estimations of direct impacts from industry circularity/material efficiency measures on emissions mitigation.

Future iterations of model development allows for more advanced solutions, we would aim for up to three high-level CE measures: reduce (Reduce (R2)), prolong (Repair (R4), Re-use (R3) and Refurbish (R5)), recycle (Recycle (R8)) (prolong might implicitly contain repair, reuse and refurbish strategies).

CE strategies and connection with climate change mitigation.

The extended MIMOSA model considers direct emissions mitigation impacts from CE actions in the industry sector.

Future iterations aim to consider emissions impacts through a breakdown of factors: carbon intensity, energy efficiency and/or material activity factors might be introduced which might follow the following cascading logic: $Emissions={\frac {Emissions}{Energy}}*{\frac {Energy}{Material}}*Material$

Insights for Analytical Framework

While CE-extended MIMOSA model does not aim at producing new results, it demonstrates key dynamics on a global level which can be further discussed and calibrated by scholars (esp. modelers) and practitioners (esp. policymakers): They may serve as a discussion starter for mechanisms that more detailed models should consider. At the same time, as more detailed models produce new data on CE impacts, the model can be recalibrated. As such it serves as a stylized toy model on a global level.

The CE-extended MIMOSA model illustrates through simple global cost-benefit modeling that CE action may be a relevant additional factor to achieve climate mitigation targets over time. If MAC curves are implemented as outlined above, the model will consider cost-optimal macro-economic dynamics across industry and non-industry sectors as well as CE and non-CE mitigation actions in the industry sector particularly. Consideration of global regions aims at indicating differing (resource) growth dynamics across developed and emerging/developing economies. Such dynamics may further influence the potential for CE strategies such as recycling (Recycle (R8)).

Future iterations of the model are motivated by additional assumptions for dynamics related to energy (separating energy use as an intermediary step in emissions calculations), CE (modelling more nuanced impact of CE measures by distinguishing demand reduction, lifetime prolongation, and recycling), material (building a simple material model linked to energy and/or emissions), sectors (modelling of other sectors) and geographies (modelling of more granular regions).

Refinement, Integration, Future Development

Refinement process

• Iterative evaluation of feasibility and development towards envisioned extensions and dynamics, depending on model complexity and data availability
• Results from bottom-up models might also be used to refine the data inputs for the model (if there is limited global data available in the literature, we might build model extensions based on sensitivity analyses of input ranges – which can then be refined by new data points)

Integration

Building on model results supplied to the CIRCOMOD data hub, model parameters may be recalibrated.

Future features of the model

In future iterations, depending on data availability, we aim to decompose the existing model's mitigation component to include energy and material dynamics as well as further sector dynamics (e.g., transport and buildings).

The energy component aims to dissect the emissions mitigation actions of the existing model by distinguishing between two traditional factors that determine emissions: energy use and energy carbon intensity.

A more advanced approach would be to develop a simple material model, where we introduce a material variable which is linked to emissions or energy and introduce high-level CE measures. Starting with, and potentially limited to, industry, we would introduce a generic material (e.g., calibrated to steel/metals, cement and plastics as the major material types for simplicity), model primary/secondary production and in-use lifetime for materials (potentially distinguishing long- versus short-lived stocks), and have three high-level CE measures (reduce, prolong, recycle).