| Variable..Chemical.Name. | Coefficient |
|---|---|
| Trifluoroethylene | 7.52 |
| PFC-14 (Perfluoromethane) | 2.84 |
| Chlorodifluoromethane | 2.29 |
| HFC-152a | 2.27 |
| Perfluoro(methylcyclopropane) | 2.24 |
| HFC-1132a; VF2 | 2.22 |
| HFC-32 | 2.19 |
| HFC-134a | 2.07 |
| 1H,10H-Perfluorodecane | 2.05 |
| Perfluorocyclobutane | 2.01 |
Model
Top Chemicals in the Model
This study’s linear regression model aims to quantify the relationship between chemical concentrations and greenhouse gas emissions. The regression equation contains a similar generalized form:
\[ \log(1 + \text{Emissions}_{\mathrm{mt}}) = \beta_0 + \sum_{j} \beta_j \cdot \text{Chemical}_j + \sum_{k} \gamma_k \cdot \text{Facility}_k + \sum_{l} \delta_l \cdot \text{Year}_l + \epsilon \]
Where B_o is the constant term, which represents the baseline level of emissions when all chemicals’ concentrations are zero. In other words, B_o provides the starting point of emissions, independent of any chemical’s influence. The coefficients (B_1, B_2, etc.) represent the strength of the relationship between each chemical’s concentration and the log-transformed emissions.
The fitted regression model identified Trifluoroethylene, PFC-14 (Perfluoromethane), and Chlorodifluoromethane as the most influential chemicals contributing to increased log-transformed emissions, with estimated coefficients of 7.52, 2.84, and 2.29, respectively. These values indicate a strong association between the presence of these chemicals and higher reported greenhouse gas emissions at the facility level.