Contributions to world emissions

These four derived variables report each region’s sector, household, government or total CO2-equivalent emissions as a percentage of world emissions from all gases. Each output has a region dimension. Gas members are summed before division; the outputs have no gas dimension.

The definitions are opt in and are available in the modified Python G-Cubed 5.0.2.0 source for build 200* models with the required emissions rows. Check the installed package catalogue when using an older installation. Regions and gases come from the selected model’s native sets.

The separate gas emissions-share variables measure the composition of an activity or total by gas. For example, EMRSHARE(CH4,USA) measures methane’s share of US emissions; REGION_EMISSIONS_SHARE_OF_WORLD(USA) measures the US contribution to world emissions across all gases. The ten gas-composition definitions retain their existing meanings.

Variables and formulas

All source quantities below are published physical mmt of CO2-equivalent emissions. For each projection year and region r, sum over every native gas member g:

Prefix	Numerator	Level calculation
SECTOR_EMISSIONS_SHARE_OF_WORLD	Total emissions of all sectors in the region	100 * sum_g(EMS(g,r)) / EMCW()
HOUSEHOLD_EMISSIONS_SHARE_OF_WORLD	Total household emissions in the region	100 * sum_g(EMH(g,r)) / EMCW()
GOVERNMENT_EMISSIONS_SHARE_OF_WORLD	Total government emissions in the region	100 * sum_g(EMG(g,r)) / EMCW()
REGION_EMISSIONS_SHARE_OF_WORLD	Total emissions in the region	100 * sum_g(EMR(g,r)) / EMCW()

Each prefix emits PREFIX(region) with ordered output set regions. The sector contribution combines all sectors in that region; it does not emit an individual-sector dimension.

EMCW() is the native scalar world total across all gases and regions. It has no gas or region dimension. The denominator is the native row for the same year, rather than a derived total repeated by region. The calculation validates that this world total agrees with the sum of published physical EMR(gas,region) values over all gases and regions. Inconsistent world accounting fails explicitly. Use accounting-consistent projections for NAB/NABb reporting.

With a positive, consistent world total, the regional contributions sum to 100%. Sector, household and government contributions retain their named source accounting; each group by itself need not sum to 100%. Summing GDP-normalised database values or averaging regional percentages would give the wrong contribution.

Levels, deviations and zero emissions

Levels have units percent. Deviations are the scenario share less the baseline share, with units percentage point deviation. A baseline contribution of 12% and a scenario contribution of 9% give a deviation of −3 percentage points. Each share uses its own world’s emissions total.

A zero numerator with a positive world total gives a zero contribution. When the world total is zero, contributions are undefined and reported as NaN; a deviation involving an undefined share is also NaN.

The definitions share the provenance handling used by the gas-composition variables. Logged-zero preparation can leave positive numerical placeholders for physically absent emissions. Restoration requires original zero observations at calibration and the preparation anchor and a match to the known published floor. True zeros and genuinely small positive emissions remain distinct. Preserve the original database and calibration information retained by ordinary saved projection objects when reusing projections; detached logged snapshots without this provenance fail validation. Calculations preserve the source projections and model inputs.

Required source rows must be complete, finite and nonnegative, use published physical mmt units and match the model’s native domain metadata. All four definitions also require EMR for the world-accounting check. A sector, household or government numerator must not exceed its region’s all-gas EMR total, allowing a relative tolerance of 1e-10 with no absolute cutoff. Missing or incompatible inputs, inconsistent world totals and overflowing totals fail explicitly.

Select contributions in setup and reports

Add the desired prefixes to a setup chartpack’s derivations list, then select regions. This setup fragment creates a chart of regional contributions to world emissions:

chartpack:
  file_name: emission-contributions.yaml
  title: Contributions to world CO2-equivalent emissions
  derivations:
    - REGION_EMISSIONS_SHARE_OF_WORLD
  charts:
    - title: Regional contributions to world emissions
      variable_prefix: REGION_EMISSIONS_SHARE_OF_WORLD
      selectors:
        - dimension: regions
          members: all

The setup instructions enable calculation in the maintained run workflow. A standalone chartpack selects rows from supplied reporting data; its series selection alone does not enable derivations. See Generate chartpacks for the complete setup format.

From Python, construct the definitions once and use them for levels and scenario deviations:

from gcubed.derivations import create_derivations

derivations = create_derivations([
    "SECTOR_EMISSIONS_SHARE_OF_WORLD",
    "HOUSEHOLD_EMISSIONS_SHARE_OF_WORLD",
    "GOVERNMENT_EMISSIONS_SHARE_OF_WORLD",
    "REGION_EMISSIONS_SHARE_OF_WORLD",
])
levels = derivations.projections(baseline_projections)
changes = derivations.deviations(scenario_projections, baseline_projections)

Pass the same collection to the runner and reporting when generating complete reports, as shown in the Python recipe.

Migrate the removed definitions

The legacy derived prefixes EMCOW and EMISSION_SHARES have been removed from the package registry. Update explicit Python callers, setup derivations lists and chartpack series that select them. There are no compatibility aliases; requesting either removed prefix fails resolution.

For regional contributions, select REGION_EMISSIONS_SHARE_OF_WORLD. This includes all gases in CO2-equivalent terms, whereas the legacy EMISSION_SHARES calculation used CO2 only. Recheck labels and interpretation when migrating a report. For a world total, select the native EMCW() model row directly rather than requesting the removed EMCOW derivation. The native scalar EMCW() remains part of the model and is the denominator for these four contribution definitions.