Page 95 - Read Online

Yu et al. Carbon Footprints 2025, 4, 17 https://dx.doi.org/10.20517/cf.2025.12 Page 11 of 23

Overall evaluation
Figure 4 presents a radar chart that offers a qualitative overview of five widely used traditional emission
models (COPERT, HBEFA, MOVES, CMEM, and VT-Micro) evaluated across five key criteria: Data
Requirements, Accuracy, Complexity, Transparency, and Transferability. The scoring values, ranging from
1 to 5, are primarily derived from accumulated domain knowledge and general experience reported in the
literature, rather than from a standardized quantitative evaluation framework.

The radar chart clearly illustrates the inherent trade-offs among these models. Microscopic models such as
CMEM and VT-Micro tend to achieve higher scores in Accuracy and Transparency, reflecting their detailed
representation of vehicle behavior and model structure. However, these models require more granular and
extensive input data, as evidenced by their higher Data Requirements scores. In contrast, macroscopic
models like COPERT and HBEFA demonstrate strengths in lower data demands and reduced Complexity,
making them more suitable for applications with limited data availability or computational resources.
Nevertheless, this often comes at the expense of reduced accuracy and diminished capacity to capture
localized emission variability. The model MOVES offers a balanced performance, with moderate scores
across the evaluated criteria.

Given the diverse application scenarios and varying data complexities involved in deploying each model,
there is currently no comprehensive scientific study that rigorously compares these models under uniform
conditions. Differences in modeling assumptions, input data quality, geographic context, and computational
methods pose significant challenges to direct benchmarking. To advance the field, future research efforts
should prioritize the development of standardized evaluation frameworks and the establishment of shared
benchmark datasets.

DATA-DRIVEN EMISSION MODELS
Data-driven emission models are computational frameworks that leverage large amounts of real-world data
to predict, analyze, or optimize pollutant emissions. These models extract patterns from historical or
real-time datasets using techniques such as machine learning, statistical analysis, and data mining,
establishing complex, often nonlinear relationships between emissions and influencing factors.

Despite increasing interest, systematic reviews of data-driven emission models remain limited. For example,
[63]
Zhang et al. categorizes data-driven models in their review of energy consumption models for new energy
vehicles, dividing them into two categories based on methodology: machine learning and neural networks.
[7]
Zhong et al. , on the other hand, classified them based on data sources into bench test data and on-road
measurement data. This paper adopts a classification framework based on the modeling object and the
machine learning approach employed.

Microscopic models
Although this paper focuses on the quantification of emissions at the traffic segment or road network level,
microscopic data-driven models are increasingly used for mesoscopic and even macroscopic
applications . These models rely on high-resolution input data, such as bench test results, OBD data, or
[64]
PEMS records, to learn the relationship between vehicle operating states and emission rates. Common input
features include VSP, speed, acceleration, and sometimes engine parameters like RPM and torque. Machine
learning models are trained to map these features to instantaneous emission rates .
[39]

Statistical regression models
Various regression and classification algorithms have been employed in emission prediction, including