DRIVE v1.0: a data-driven framework to estimate road transport emissions and temporal profiles
Daniel Kühbacher, Jia Chen, Patrick Aigner, Mario Ilic, Ingrid Super, Hugo Denier van der Gon · 2025 · 2 citationsRead the paper
Abstract. Traffic in urban areas is an important source of greenhouse gas (GHG) and air pollutant emissions. Estimating traffic-related emissions is therefore a key component in compiling a city emission inventory. Inventories are fundamental for understanding, monitoring, managing, and mitigating local pollutant emissions. We present DRIVE v1.0, a data-driven framework to calculate road transport emissions based on a multi-modal macroscopic traffic model, vehicle class-specific traffic counting data from more than a hundred counting stations, and HBEFA emission factors. DRIVE introduces a novel approach for estimating traffic emissions with vehicle-specific temporal profiles in hourly resolution. In addition, we use traffic counting data to estimate the uncertainty of traffic activity and the resulting emission estimates at different temporal aggregation levels and with road link resolution. The framework was applied to the City of Munich, covering an area of 311 km2 and accounting for GHGs (CO2, CH4) and air pollutants (PM, CO, NOx). It captures irregular events such as COVID lockdowns and holiday periods well and is suitable for use in near real-time applications. Emission estimates for 2019–2022 are presented and differences in city totals and spatial distribution compared to the official municipal reported and national and European downscaled inventories are examined.
What the full text shows
Working prototype- Estimated total vehicle kilometers traveled (VKT) in Munich showed a decrease during the pandemic followed by a partial recovery. — 12.7% decrease in 2020, 1.5% increase in 2021, and 3.5% increase in 2022 (City of Munich, 2019–2022)
- The framework's 2019 emission estimates were lower than the city's official assessment but differed significantly from downscaled national and European inventories. — 6.2% lower than city assessment; UBA national inventory was 47% higher and TNO European inventory was 24% lower for CO2 (City of Munich, 2019)
- Comparison of VKT for 2019 against the RKU (municipal) method showed differences based on the inclusion of minor roads and temporal scaling. — 5.2% lower VKT for heavy goods traffic, 7% higher for other vehicle classes, and 6.4% higher total VKT (City of Munich, 2019)
- Road link level emission uncertainty varies by pollutant and temporal aggregation, with the lowest uncertainty typically at the daily level. — Daily uncertainty: NOx 31.9%, CO 33.8%, CO2 28.0% (Road link level resolution)
- Traffic fluidity increased over the study period, with a higher proportion of travel occurring under free-flow conditions. — Free-flow VKT increased from 53% in 2019 to 59% in 2020 (City of Munich, 2019–2020)
- Setup
- Applied to the City of Munich (311 km2) using 116 traffic counters (82 municipal, 34 BASt) and HBEFA 4.2 emission factors for 2019–2022.
- How
- The DRIVE v1.0 framework combines a macroscopic traffic demand model (VISUM) with dynamic hourly data from 116 traffic counting stations to scale traffic activity. It applies HBEFA 4.2 emission factors based on vehicle class, road type, gradient, and a volume-capacity ratio to estimate hourly hot and cold start emissions.
- Limitations
- Uncertainty analysis excludes cold start emissions for CO, likely underestimating its uncertainty. · Relies on statistical averages from HBEFA for fleet composition rather than city-specific powertrain data. · Volume-capacity ratio for congestion estimation is less accurate at intersections than on road links. · Does not explicitly account for congestion effects like queues and spillbacks.
Summarized by Gemma 4 from the open-access full text · 9 Oct 2026. Read it