CO2 transport and dispersion processes are modeled using the WRF-Chem modeling system (Grell et al. 2005), with emission input based on Hestia (2012) and meteorological input based on the 32-km resolution North American Regional Reanalysis (NARR). To constrain the model error, available observations such as surface and upper-air observations from the World Meteorological Organization (WMO) are assimilated using the Four Dimensional Data Assimilation (FDDA) technique developed and implemented into WRF by Penn State (Deng et al. 2009). Ongoing work includes assimilating additional meteorological observations available to the INFLUX project such as the wind profile from the local HALO lidar deployed by NOAA Earth System Research Laboratory Chemical Sciences Division (http://www.esrl.noaa.gov/csd/groups/csd3/measurements/influx/), and the Aircraft Communications Addressing and Reporting System (ACARS) commercial aircraft observations.
The current INFLUX WRF configuration consists three nested grids with 9-/3-/1-km horizontal resolutions, and 60 vertical model layers, with the first model layer at about 7 m AGL and with 24 model layers below 1.5 km AGL (Fig. 1). The model physics includes the MYNN PBL scheme, Noah land surface scheme, RRTM longwave and Dudhia shortwave radiation schemes, and WSM-5 microphysical scheme. The Kain-Fritsch cumulus parametrization scheme is used on the 9-km grid. For FDDA, both analysis nudging and observational nudging are used to assimilate observations (Deng et al. 2009). As shown in Table 1, WRF simulations covering a year-long period with FDDA assimilating available standard WMO surface and upper-air observations has been completed. Model solutions can be viewed at http://www.meteo.psu.edu/~ngafs1/indi_v3.5.1_longrun_rerun/ for the September 2012-April 2013 time period and at http://www.meteo.psu.edu/~axd157/indi_v3.5.1_longrun/ for the May-Oct 2013 time period.
The WRF model errors are estimated by computing the error statistics between the model values and the observed values of the meteorological variables including wind speed, wind direction, temperature and specific humidity. Model snapshots at hourly output intervals are compared with the observations near the same time within a time window of 0.5 hour (15 minutes on each side of the hour). To compute the error statistics, model values are interpolated to the observation locations, and then the model and obs pairs are used to compute the mean absolute error (MAE), mean error (ME) and root mean square error (RMSE) of the meteorological variables. The model error statistics for the September 2012-April 2013 time period can be found at http://www.meteo.psu.edu/~ngafs1/indi_v3.5.1_longrun_rerun/stats/stats.html and the model error statistics for the May-Oct 2013 time period can be found at http://www.meteo.psu.edu/~axd157/indi_v3.5.1_longrun/stats/stats.html
- Deng, A., D.R. Stauffer, B.J. Gaudet, J. Dudhia, J. Hacker, C. Bruyere, W. Wu, F. Vandenberghe, Y. Liu and A. Bourgeois, 2009: Update on WRF-ARW end-to-end multi-scale FDDA system, 10th Annual WRF Users’ Workshop, Boulder, CO, June 23-26, 14 pp (Available on the web at: http://www.mmm.ucar.edu/wrf/users/workshops/WS2009/abstracts/1-09.pdf).
- Grell, G. A., S. E. Peckhama, R. Schmitzc, S. A. McKeenb, G. Frostb, W. C. Skamarockd, B. Edere, 2005: Fully coupled “online” chemistry within the WRF model, Atmos. Environ. 39, 6957-6975, doi:10.1016/j.atmosenv.2005.04.027.
FAIR USE POLICY
Permission to download this data does not grant permission to use the information contained in the data for publication or any commercial use.
Please do not distribute.
To request permission to access data, please send an email to email@example.com describing your plans for use.