Once we have developed modules, methods and tools in the project, they can be found on this page.
The detailed descriptions of the methodologies respond to the widespread need in the modelling community to share methodologies and input data for the further improvement of integrated assessment and energy‐economy modelling. The toolbox includes new model components, mathematical formulae, algorithmic approaches, examples of model code, and generic input datasets and instructions for implementation.
Datasets developed in the NAVIGATE project
Inequality data set
This novel data set includes data on household characteristics on income, expenditures, savings rates, educational attainment, and notably, expenditure shares on energy (transportation and buildings) and food consumption. The data set provides consistent information at the household income decile level, which can be used in IAMs and other models to calibrated income distributions in the models, and has already been used by several models.
Decile-based input dataset at the iso3 level for latest available survey data to calibrate initial year distributional variables.
Data set elements
Within NAVIGATE, a standardized input data template has been defined and data has been collected using household surveys from a number of developed and OECE countries. Available countries (for the latest year, between 2008 and 2017, available), include India, Brazil, France, South Africa, United States of America, China, Mexico, as well as 26 EU member states based on a common survey.
The data is all structured around income deciles. Units of observations for the deciles are households. The OECD modified equivalence scale has been applied to compute household income per person. (1 plus 0.5 per adult (older than 14 years) and 0.3 per child (14 or less years)).
Available variables (all for deciles D1 to D10):
Energy for Transportation
Energy for Housing
Education of the household head
Average savings rate
Income by type (labour, capital, transfers, other)
Wealth share (where available)
The energy expenditure shares for housing and transportation are key variables for IAMs, and here the expenditure shares across countries, which show a strong regressive pattern for residential energy consumption while for transportation it appears in most countries to be progressive.
Link to code/data
How to use the data/module
Load the CSV file “deciles_data.csv” in your model at the ISO3 level. Further data description in the process_survey_data.R file.
Probabilistic, long-term (2100) marginal abatement cost curves for CH4 and N2O emissions (2022)
This dataset contains long-term (up to 2100) CH4 and N2O Marginal Abatement Cost (MAC) curves for all major emission sources, with “optimistic”, default and “pessimistic” assumptions on mitigation potentials (i.e. high, medium, low reduction potentials, respectively).
The MACs have been developed and assessed with the IMAGE integrated assessment model. This will be described in an upcoming paper: “Uncertainty in non-CO2 greenhouse gas mitigation: Make-or-break for global climate policy feasibility” (Harmsen et al.). The component-based construction of the MACs is based on “Long-term marginal abatement cost curves of non-CO2 greenhouse gases” Mathijs Harmsen et al (Environmental Science & Policy, 2019), https://doi.org/10.1016/j.envsci.2019.05.013
The MACs in the dataset are an update of the MACs from Harmsen et al., 2019, incorporating all literature and almost all mitigation measures from that study, but complemented with insights from recent literature (including new measures and new measure-specific studies). So note: values from the 2019 study vary somewhat from the default MAC in the new dataset.
The MACs have been developed for all major non-CO2 emission sources, but in most detail for agricultural sources, since 1) these are hardest-to-abate in mitigation cases, 2) have high uncertainty and 3) these could be constructed fully bottom up, based on Harmsen et al. (2019). The agriculture MACs have been built-up from quantitative components. In a Monte Carlo simulation, these input parameters were varied to determine lower and upper bounds for the overall mitigation potentials. The non-agriculture MAC ranges have been determined by varying the maximum reduction potentials, also based on recent literature.
Module elements (data description)
The datafile (“Data_MAC_CH4N2O_Harmsen et al_PBL”) contains relative emission reductions for the different CH4 and N2O emission sources at different marginal cost levels for the period 2015-2100 (not all intermediate years are provided, but values between subsequently provided years can be linearly interpolated). Further specifications are provided in the information sheet in the dataset document. Note (!) that the MAC curves are baseline-independent; Values represent relative reductions compared to the global average emission factor in 2015 for the emission source concerned. Negative values represent a higher (regional) emission factor than the global average in 2015. So, this occurs in high emission-intensity regions, where prompt lowering of emission factors to the global average is unlikely. Negative values are provided for several non-agricultural sources. For agricultural sectors, regional differences in emission-intensities obviously also exist, but this is not reflected in the MACs’ reduction potentials (due to a different construction approach). Teams are therefore also advised to avoid rapid changes in relative emission reductions, by accounting for inertia (see below).
In the long term and at very high prices, the regional reduction potentials converge, unless biophysical differences (mainly climatic) are known to lead to differences in long-term mitigation potentials (e.g. one can’t introduce Holstein cows, with relatively favourable production/methane ratios, in hotter regions).
Link to code/data
How to use the module
In order to implement the MACs, model teams need to know: global, regional and source-specific emission factors in time (to compensate for reduction in the baseline and simulate regional convergence at high prices). See emission source definition and region specification below. Teams are also advised to avoid rapid changes in relative emission reductions, by accounting for inertia, e.g. by introducing a maximum yearly allowed change in relative reduction (as is done in IMAGE).
For questions, please contact: Mathijs Harmsen (firstname.lastname@example.org).
The study will be published open access under a Creative Commons licence. The data can be freely used by all parties.
Harmsen et al. (in preparation)
Mathijs Harmsen, Charlotte Tabak, Lena Höglund-Isaksson, Pallav Purohit, Detlef van Vuuren Uncertainty in non-CO2 greenhouse gas mitigation: Make-or-break for global climate policy feasibility
Harmsen et al., 2019
Mathijs J.H.M. Harmsen, Detlef P. van Vuuren, Dali R. Nayak, Andries F. Hof, Lena H€oglund-Isaksson, Paul L. Lucas, Jens B. Nielsen, Pete Smith, Elke Stehfest, Long-term marginal abatement cost curves of non-CO2 greenhouse gases, Environ. Sci. Policy 99 (2019) 136e149. https://doi.org/10.1016/j.envsci.2019.05.013
Long-term marginal abatement cost curves of non-CO2 greenhouse gases (2019)
This dataset represents long-term marginal abatement cost (MAC) curves of all major emission sources of non-CO2 greenhouse gases (GHGs); methane (CH4), nitrous oxide (N2O) and fluorinated gases (HFCs, PFCs and SF6). The work is based on existing short-term MAC curve datasets and recent literature on individual mitigation measures. The data represent a comprehensive set of MAC curves, covering all major non-CO2 emission sources for 26 aggregated world regions. They are suitable for long-term global mitigation scenario development, as dynamical elements (technological progress, removal of implementation barriers) are included. The data is related to the research article Harmsen et al. (2019) “Long-term marginal abatement cost curves of non-CO2 greenhouse gases” (https://doi.org/10.1016/j.envsci.2019.05.013)
The datasets contain CH4 and N2O (Data_MAC_CH4N2O_Harmsen et al._PBL) and fluorinated gas (Data_MAC_F-gases_Harmsen et al._PBL) marginal abatement cost (MAC) curves for all major global emission sources. Values represent relative emission reductions for the different emission sources at different marginal cost levels for the period 2015e2100 (not all intermediate years are provided, but values between subsequently provided years can be linearly interpolated). Two sets are made available: 1) One baseline-independent set with relative reductions compared to the global average emission factor in 2015 for the emission source concerned. Negative values represent a higher emission factor than the global average in 2015. 2) One set compatible with the IMAGE SSP2 baseline scenario (with “SSP200 in the name of the sheet). Source-specific emission reductions in SSP2 are deducted from the reductions in the baseline-independent MACs. Implementation costs are provided in (2005/2010) $/tonne of C equivalents, assuming the use of the AR4 100 yr GWP potential.
The MAC curves represent the combined reduction potential of all relevant mitigation measures at specific marginal costs for a specific emission source and country or region. In order to be relevant for long term climate policy projections, they account for future changes in reduction potential and costs, due to 1) technological learning and 2) removal of implementation barriers. The MAC curves developed in this study are based on a combination of existing datasets and an assessment of individual mitigation options described in literature.
Link to code/data
The data files can be downloaded here: https://doi.org/10.1016/j.dib.2019.104334
How to use the dataset
For integrated assessment modelling teams, the recommendation is to use the first dataset: baseline-independent set with relative reductions compared to the global average emission factor in 2015 for the emission source concerned. Note that, when implementing these MACs in an IAM, relative emission reductions as represented in the dataset cannot simply be deducted from the baseline emissions. The model should take into account emission reductions already taking place in the baseline.
The study has been published open access under a Creative Commons licence. The data can be freely used by all parties.
Mathijs J.H.M. Harmsen, Detlef P. van Vuuren, Dali R. Nayak, Andries F. Hof, Lena H€oglund-Isaksson, Paul L. Lucas, Jens B. Nielsen, Pete Smith, Elke Stehfest (2019) Long-term marginal abatement cost curves of non-CO2 greenhouse gases, Environ. Sci. Policy 99 136e149. https://doi.org/10.1016/j.envsci.2019.05.013
Mathijs J.H.M. Harmsen, D.P. van Vuuren, D.R. Nayak, A.F. Hof, L. Höglund-Isaksson, P.L. Lucas, J.B. Nielsen, P. Smith, E. Stehfest (2019) Data for long-term marginal abatement cost curves of non-CO2 greenhouse gases, Data Brief, 25, p. 104334, https://doi.org/10.1016/j.dib.2019.104334
Peatland maps and GHG emission factors
The data set includes maps of degraded (~46 Mha globally) and intact peatland (~375 Mha globally) for the year 2015. The spatial resolution is 0.5 degree. The data set also includes IPCC wetland GHG emission factors for degraded and rewetted peatlands.
Link to data
Supplementary data in Humpenöder et al 2020 ERL
How to use the data set
The data set can be used to initialize intact and degraded (drained) peatland area in land use modules of IAMs. In MAgPIE, the future peatland dynamics depend on a peatland scaling factor (see methods and SI in Humpenöder et al 2020 ERL). GHG emissions from degraded peatlands and GHG emission savings from restored (rewetted) peatlands can be estimated by multiplication with the corresponding wetland GHG emission factors. Details on the usage of the data set in MAgPIE: https://github.com/magpiemodel/magpie/tree/develop/modules/58_peatland/on
The study has been published open access under a Creative Commons license. The data can be freely used by all parties.
Humpenöder F, Karstens K, Lotze-Campen H, Leifeld J, Menichetti L, Barthelmes A and Popp A 2020 Peatland protection and restoration are key for climate change mitigation Environ. Res. Lett. 15 104093. https://doi.org/10.1088/1748-9326/abae2a
Dataset on AFOLU mitigation and biomass potentials
This dataset provides a model emulation (so called “Lookup-Table”) of the GLOBIOM-G4M models with respect to greenhouse gas (GHG) emission reduction potentials from agriculture, forestry and other land use (AFOLU) and land based biomass potentials for bioenergy. The dataset represents a two dimensional scenario matrix combining different carbon price and biomass price trajectories for the Shared Socio-economic Pathway 2 (SSP2) and can be used in other models e.g. energy system models, to develop climate mitigation pathways that explicitly consider impacts/potentials from the land use sector.
Link to code/data
The GLOBIOM-G4M Lookup-Table can be found here.
Detailed information on the GLOBIOM-G4M Lookup-Table structure and results are presented our paper ‘Land based climate change mitigation potentials within the agenda for sustainable development’ and the accompanying supplementary material.
Information on the GLOBIOM model can be found here.
How to use the dataset
The Lookup-Table represents a combination of 7 biomass price and 12 GHG price trajectories that have been quantified in GLOBIOM-G4M, yielding in total 84 scenarios. The biomass price for bioenergy production (US$/GJ) and carbon price (US$/tCO2eq) increase linearly from 2020 onwards and reach their maximum in 2100. Prices range from 0 – 60 US$/GJ (BIO00, BIO03, BIO05, BIO08, BIO13, BIO30, and BIO60) and 0 – 3000 US$/tCO2eq (GHG000, GHG010, GHG020, GHG050, GHG100, GHG200, GHG400, GHG600, GHG1000, GHG1500, GHG2000, and GHG3000). This approach allows to quantify supply functions where the supplied biomass quantity available for bioenergy is a function of the biomass price and conditional on a GHG price. Vice versa we quantify the cost-efficient AFOLU mitigation potential in the form of a marginal abatement cost curves (MACC) conditional on the biomass demand where the emission reduction is a function of the GHG price converted through global warming potential of the non-CO2 gases to cover also methane (CH4) and nitrous oxide (N2O) emissions in addition to carbon dioxide (CO2).
The biomass supply curves and MACCs are highly interdependent, for instance: a biomass price remunerating forest harvest for bioenergy may encourage additional afforestation. This will increase the carbon sink of the forest while at the same time providing more biomass for bioenergy production. The Lookup-Table represents a GLOBIOM-G4M model emulation and provides a comprehensive and detailed response surface for the land use sector that can also be used in other models to explicitly consider dynamics and interlinkages between biomass use and AFOLU emissions but also other important land use related indicators.
Detailed information on the SDG scenario dimensions, regional aggregation and output variables can be found here.
License information and publication
Frank S, Gusti M, Havlík P, Lauri P, DiFulvio F, Forsell N, Hasegawa T, Krisztin T, Palazzo A, and Valin H (2021). Land based climate change mitigation potentials within the agenda for sustainable development. Environmental Research Letters. Volume 16(2)024006.
This module includes updated SSP GDP per capita scenarios (covering the Covid shock), related PPP/MER conversion rates, and a set of structural change scenarios. The latter is represented by sectoral shares on employment, value added and final energy use. The sectoral resolution contains agriculture, manufacturing and services. The development of these key variables of economic activity is projected until 2050 and is available for each the five SSP scenarios. GDP per capita scenarios are available until 2100. The data set also includes historical data (from 1998 onwards) for GDP, population and sectoral shares.
The scenarios are available on a country and region level. A more in-depth description of the scenarios is available here.
Link to code/data
The data set is provided in this excel file.
How to use the scenarios
The structural change scenarios are directly linked to the updated set of SSP GDP scenarios. Updated GDP scenarios can be used as basic driver and the structural change scenarios as an additional driver in all mitigation and impact analyses.
Leimbach, M., Marcolino, M., Koch, J. (2021). Structural change scenarios within the SSP framework. Futures (submitted).
Koch, J. and Leimbach, M. (2021). Update of SSP GDP projections: capturing recent changes in national accounting, PPP conversion and Covid 19 impacts. Working paper.
Modules and tools
Modules and tools developed in the NAVIGATE project
Technology diffusion indicators and plots
Link to code
IAM diagnostics tool – Diagnostic indicators and graphs
This tool / short code, written in Python, contains 1) calculations of 6 key diagnostic indicators of the behaviour of integrated assessment models (IAMs) based on IAM output (as described in Harmsen et al., 2021) and 2) code to create graphs that visualize the diagnostic results for multiple IAMs. The tool is a useful asset for IAM modellers that want to perform diagnostic analyses with their model or with multiple models. As described in Harmsen et al.:
“Integrated assessment models (IAMs) form a prime tool in informing about climate mitigation strategies. Diagnostic indicators that allow comparison across these models can help describe and explain differences in model projections. This increases transparency and comparability. Earlier, the IAM community has developed an approach to diagnose models (Kriegler et al., 2015). Here we build on this, by proposing a selected set of well-defined indicators as a community standard, to systematically and routinely assess IAM behaviour, similar to metrics used for other modeling communities such as climate models. These indicators are the relative abatement index, emission reduction type index, inertia timescale, fossil fuel reduction, transformation index and cost per abatement value. We apply the approach to 17 IAMs, assessing both older as well as their latest versions, as applied in the IPCC 6th Assessment Report. The study shows that the approach can be easily applied and used to indentify key differences between models and model versions. Moreover, we demonstrate that this comparison helps to link model behavior to model characteristics and assumptions. We show that together, the set of six indicators can provide useful indication of the main traits of the model and can roughly indicate the general model behavior.”
The code contains calculations of the 6 key indicators and scripts to create graphs based on the indicator output.
The tool also includes the dataset used in the Harmsen et al. paper.
Link to code/data
Link to the code and data (By: Kaj-Ivar van der Wijst, 2021)
How to use the tool
The tool also includes the dataset used in the Harmsen et al. paper and can thus be run directly without external input. However, if the code is used to add diagnostic results for new models or new model versions, it is necessary to run the diagnostic scenarios with these models first and to add the scenario results to the dataset in the tool. The diagnostic scenarios have a very simple setup. Models are confronted with a stylized carbon price profile. The diagnostic indicators are based on the model outcomes in 2050. See the Harmsen et al. paper for a description of the scenarios and indicators, or contact Mathijs.email@example.com.
The underlying study has been published open access under a Creative Commons licence. The data can be freely used by all parties.
Harmsen, M. et al (2021) Integrated assessment model diagnostics: key indicators and model evolution. Environ. Res. Lett. 16 054046. https://doi.org/10.1088/1748-9326/abf964
Link to the code and data (By: Kaj-Ivar van der Wijst, 2021), https://doi.org/10.5281/zenodo.5727623
The pyam package for IAM scenario analysis & visualization
This package provides a suite of tools and functions for analyzing and visualizing input data (i.e., assumptions, parametrization) and results (model output) of integrated-assessment scenarios, energy systems analysis, and sectoral studies.
The package is based on the time series data format developed by the Integrated Assessment Modeling Consortium (IAMC), but it supports additional features such as sub-annual time resolution.
Manuscript in Open Research Europe: https://doi.org/10.12688/openreseurope.13633.1
GitHub repository: https://github.com/IAMconsortium/pyam
Community forum: https://pyam.groups.io
Aviation Integrated Model
The Aviation Integrated Model (AIM) is a global aviation systems model which simulates interactions between passengers, airlines, airports and other system actors into the future, with the goal of providing insight into how policy levers and other projected system changes will affect aviation’s externalities and economic impacts. The model was originally developed in 2006-2009 with UK research council funding (e.g. Reynolds et al., 2007; Dray et al. 2014), and was updated as part of the ACCLAIM project (2015-2018) between University College London, Imperial College and Southampton University (e.g. Dray et al., 2019), with additional input from MIT regarding electric aircraft (e.g. Schäfer et al., 2018). The model is open-source, with code, documentation and a simplified version of model databases which omit confidential data available from the UCL Air Transportation Systems Group website (note that the website code and databases are slightly simplified from the full version used at UCL to remove confidential data).
In the context of NAVIGATE, a metamodeling approach is taken. We consider the most important factors affecting future aviation demand and emissions to be:
- Socioeconomic scenario (e.g. population, GDP, and potentially changes in attitudes to flying),
- Oil price,
- Carbon price, and
- Technology characteristics.
For each of these factors, we define a range of model inputs and carry out a grid of model runs using those inputs. More in-depth information on the model inputs can be found here.
Link to code/data
The metamodel and data needed to use it are available from the GitHub public repository: https://github.com/ODessens/NAVIGATE_T3.3
How to use the model
Currently, the model is supplied as two Python code files and a set of associated data tables. These routines also contain extensive comments on how they function and on the definition of different variables. Each model run requires two separate components. First, the data tables are read in. Second, each time the main IAM using the aviation metamodel requires aviation metrics, the interpolation model is run.
The different functions are elaborated upon in this document, containing more in-depth information about the model.
The data and program can be freely used by all parties.
Reynolds, T., Barrett, S., Dray, L., Evans, A., Köhler, M., Vera-Morales, M., Schäfer, A., Wadud, Z., Britter, R., Hallam, H., Hunsley, R., 2007. Modelling Environmental and Economic Impacts of Aviation: Introducing the Aviation Integrated Modelling Tool. In: Proceedings of the 7th AIAA Aviation Technology, Integration and Operations Conference, Belfast, 18–20 September 2007, AIAA-2007-7751;
Dray, L., Evans, A., Reynolds, T., Schäfer, A., Vera-Morales, M. and Bosbach, W., 2014. Airline fleet replacement funded by a carbon tax: an integrated assessment. Transport Policy, 34, 75-84.
Dray L., Krammer P., Doyme K., Wang B., Al Zayat K, O’Sullivan A., Schäfer A., 2019. “AIM2015: Validation and initial results from an open-source aviation systems model”, Transport Policy, 79, 93-102.
Schäfer A., Barrett, S., Doyme, K., Dray, L., Gnadt, A., Self, R., O’Sullivan, A., Synodinos, A., & Torija, A., 2018. Technological, economic and environmental prospects of all-electric aircraft. Nature Energy, 4, 160-166.
Other relevant publications are cited in this document, which contains more in-depth information.