Municipal Emissions Attribution

Disaggregate citywide CO₂ into building sectors, types, vintages, and end uses using ResStock and ComStock emissions columns.

When to Use

You have an aggregate emissions figure for a city or county and need to break it down into actionable categories: which building types are responsible, which construction eras, which end uses (HVAC vs lighting vs plug loads). This is the difference between “Atlanta emitted X tons of CO₂” and “pre-1980 single-family homes in Fulton County account for 34% of residential heating emissions.”

Pipeline

Download state ResStock/ComStock baseline
    → filter to target county
    → multiply emissions × weight
    → group by building_type / vintage / end_use
    → attribution breakdown

Prerequisites

Download Georgia ResStock baseline data per the Accessing the Data guide.

Complete Code

Step 1: Load and filter to your county

import pandas as pd

df = pd.read_csv("georgia_resstock_baseline.csv", low_memory=False)

# Filter to Fulton County (Atlanta)
# ResStock uses GISJOIN codes — Fulton County, GA = G1312100
fulton = df[df["in.county"] == "G1312100"].copy()
print(f"Buildings in Fulton County: {len(fulton):,}")

Step 2: Emissions by building type

# CO₂ emissions column (pounds, LRMER mid-case 15-year scenario)
emissions_col = "out.emissions.co2e.lrmer_mid_case_15.all_fuels.total.lb"

# Always weight — each model represents multiple real buildings
fulton["weighted_co2_lb"] = fulton[emissions_col] * fulton["weight"]

by_type = (fulton.groupby("in.geometry_building_type_recs")["weighted_co2_lb"]
           .sum()
           .sort_values(ascending=False))

# Convert pounds to metric tons
by_type_tons = by_type / 2_204.6

total_tons = by_type_tons.sum()
print(f"\nFulton County residential CO₂ by building type:")
print(f"{'Building Type':<30} {'Metric Tons':>12} {'Share':>8}")
print("-" * 52)
for btype, tons in by_type_tons.items():
    print(f"{btype:<30} {tons:>12,.0f} {tons/total_tons:>7.1%}")

Step 3: Emissions by vintage (construction era)

by_vintage = (fulton.groupby("in.vintage")["weighted_co2_lb"]
              .sum()
              .sort_index())

by_vintage_tons = by_vintage / 2_204.6

print(f"\nFulton County residential CO₂ by construction era:")
print(f"{'Vintage':<20} {'Metric Tons':>12} {'Share':>8}")
print("-" * 42)
for vintage, tons in by_vintage_tons.items():
    print(f"{vintage:<20} {tons:>12,.0f} {tons/total_tons:>7.1%}")

Step 4: Emissions by end use

ResStock provides emissions broken down by fuel and end use. To see which activities drive emissions:

end_use_cols = [c for c in fulton.columns
                if c.startswith("out.emissions.co2e.lrmer_mid_case_15.")
                and c.endswith(".lb")
                and "all_fuels" not in c
                and ".total." in c]

end_use_data = {}
for col in end_use_cols:
    # Extract fuel type from column name
    parts = col.split(".")
    fuel = parts[4]  # e.g., "electricity", "natural_gas"
    weighted = (fulton[col] * fulton["weight"]).sum() / 2_204.6
    if weighted > 0:
        end_use_data[fuel] = weighted

print(f"\nFulton County residential CO₂ by fuel type:")
for fuel, tons in sorted(end_use_data.items(), key=lambda x: -x[1]):
    print(f"  {fuel:<20} {tons:>12,.0f} metric tons ({tons/total_tons:.1%})")

Reading the Results

What “LRMER Mid Case 15” means: Long-Run Marginal Emissions Rate, mid-case grid decarbonization scenario, 15-year time horizon. This represents the emissions impact of changes in electricity demand, accounting for how the grid is expected to evolve. It’s the recommended scenario for evaluating future interventions.

Other available scenarios (substitute in the column name):

lrmer_high_re_cost_15 — pessimistic grid decarbonization
lrmer_low_re_cost_15 — optimistic grid decarbonization
aer_mid_case_1 — average emissions rate (for carbon inventory/accounting, not intervention planning)

Unit conversion: ResStock reports in pounds (lb). Divide by 2,204.6 for metric tons, or by 2,000 for short tons.

Adding Commercial Buildings

The code above uses ResStock (residential). To get the full picture, repeat with ComStock (commercial) data:

com = pd.read_csv("georgia_comstock_baseline.csv", low_memory=False)

# ComStock uses a different county column format — check your data dictionary
# Filter, weight, and group the same way

Combine residential + commercial totals for a complete county-level attribution.