Update on the Use of WRF TKE Profiles in SCIPUFF Dipole Pride 26 Field Test Comparison

Diffusion in SCIPUFF, a hazard-prediction model, is typically based on simple assumed profiles describing turbulence for equilibrium neutral and convective conditions over uniform surfaces. Within the planetary boundary layer (PBL), these profiles are functions of height normalized by the PBL height, while uniform values are assumed above the PBL. SCIPUFF's assumed profiles for describing turbulence are thus oversimplifications for many situations involving wind shear, cloud, terrain and other surface inhomogeneity. Following our WRF-SCIPUFF simulations for CAPTEX presented at the last GMU conference, we now investigate the use of WRF turbulent kinetic energy (TKE) and mixing length profiles to drive SCIPUFF dispersion model calculations for Dipole Pride 26 (DP26). Our previous results showed some benefit using WRF TKE and mixing length profiles in the SCIPUFF simulations of the CAPTEX tracer when using the MYJ and MYNN PBLs. The largest positive effects occurred in the tail of the plume and when using MYJ. The WRF V4.2 is integrated again here for DP26 with data assimilation including the special observations and a spinup period to allow cold-air drainage, afternoon upsloping or channeling by the terrain to develop over the Nevada Test Site, Yucca Flat, during November 1996. Five cases using a 9 / 3 / 1 -km nested grid configuration were chosen for testing with the MYJ and MYNN turbulence schemes. Both PBL schemes predict TKE for eddy diffusivity while the latter also includes an eddy diffusivity mass flux (EDMF) formulation and TKE advection. Compared to our 3-km CAPTEX long-range atmospheric transport and dispersion (1-3 days, 300-1100 km) simulations, our 1-km short-range DP26 atmospheric transport and dispersion simulations are on much finer temporal and spatial scales (3 h, 2-20 km). Our five cases include both northern and southern valley puff releases measured downwind along three sensor lines. The SCIPUFF results are compared to the 3-h line-maximum surface dosages. Statistical results using WRF-SCIPUFF suggest that use of TKE improves the 3-h maximum or line-integrated dosages when using the MYJ PBL. The MYNN experiments produce smaller surface dosages than the MYJ possibly due to its deeper PBLs and greater TKE / vertical mixing. Comparisons of WRF MYJ and MYNN TKE results in the literature often show MYNN producing larger and more accurate values of TKE. So the greater advantage using TKE in the WRF MYJ to drive SCIPUFF for both CAPTEX and DP26 tracer studies requires further investigation.