13.2 Emissions from a Known Location


Previous

HOME


Next

In this section we will use the measurement data to estimate the magnitude of the emissions assuming we know the location of the source. Start by retrieving the previously saved captex_control.txt and captex_setup.txt settings into the GUI menu.

  1. To simplify the example and produce results more quickly, we will only use the sampling data from the 3-hour duration collections. These occurred from 18Z September 25th to 12Z September 26th. Therefore, in the Concentration / Setup Run menu, change the run duration to 19 hours. Save to close and exit all menus.

  2. Run the model and when the simulation completes, open the Convert to DATEM utility menu and browse for the measured data input file of the 3-hour duration data. You could manually create a subset from the master file captex2_meas.txt, but for convenience, this file is already provided captex2_3hr.txt. Fill in the remaining fields as required, then as in previous sections, create the DATEM file, compute the statistics, and open the Scatter Plot. The ground-level sampling covers the same time period as the aircraft sampling data used in the some of the previous sections and the model results are quite good for these data with a ratio of mean calculated to measured of 0.98 and a correlation coefficient of 0.52.

  3. To use the measurement data to estimate emissions from the release location we need to agree about what is known and not known. In this case we assume that we know nothing about the emission rate nor the duration of the emissions, but we do know the release location. Open the Setup Run menu and start the simulation at the beginning of the meteorological data file. Change the start time from 17 to 15, and increase the run duration from 19 to 21 hours. Now open the Pollutant Emission menu and change the emission rate from 67000 to 1.0 and the hours of emission from 3.0 to 21.0. Save the changes and run the model.

  4. After the simulation completes open the Convert to DATEM menu, use the same 3-hr sampling data file, the unit conversion factor 1.0E+12, create the DATEM file and compute the statistics. The scatter plot is not needed.

  5. By using a unit emission rate we can easily compute the actual emission rate by dividing the measured values by the model computed dispersion factor. Consider that the measured concentration (M) represents the product of the atmospheric dilution (D) from the source to the receptor and the emission rate (Q):
    • M [g/m3] = D [hr/m3] Q [g/hr]

    Because the model computes the dilution factor, we can simply re-arrange the equation and adjusting for units:
    • Q [g/hr] = M [pg/m3] / ( D [hr/m3] * 1E+12 [pg/g])

    Looking at the statistical results output, the mean measured concentration is 2303.92 and the mean calculated is 0.07, which means that the estimated emission rate is 32,914 g/hr (2304/0.07). We know that the actual emission rate was 67,000 g/hr but for only a 3-hour period. Because of the stationarity in the flow field, contributions from other time periods with no emissions also contributed to the predictions at samplers with measurements, resulting in a lower emission rate being able to satisfy the solution to the above equation.

  6. It may be possible to infer the emission time variation by examining the M/D ratios by time period. The DATEM statistical program also creates an output file of the measurements and calculations paired by time and sampling location. Examining the dataA.txt file for the locations with the highest concentrations and computing the M/D ratio shows considerable variability but no clear temporal trends. The zero predictions prior to 00Z on the 26th in combination with an estimate of the transport time could be used to estimate the emission start time.

  7. Save the control and namelist files to srcfwrd_control.txt and srcfwrd_setup.txt.

Measurement data in combination with the dilution factors computed by the model were used to estimate the source term from a known source location. There is considerably more uncertainty in estimating emissions than concentrations because the precision of the result is limited by the averaging time of the measurement data collection. Small errors in model transport can result in large errors in the emission estimate because the dilution factor is in the denominator and transport errors can cause its value to approach zero.