<!DOCTYPE HTML PUBLIC "-//W3C//DTD HTML 3.2//EN">
<HTML>
<HEAD>
<META HTTP-EQUIV="Content-Type" CONTENT="text/html; charset=iso-8859-1">
<META NAME="Generator" CONTENT="MS Exchange Server version 6.5.7651.59">
<TITLE>FW: Using Diffusion TKE equation at 1 km</TITLE>
</HEAD>
<BODY>
<!-- Converted from text/plain format -->
<P><FONT SIZE=2><BR>
<BR>
Dear wrf-users:<BR>
<BR>
I am attempting to model TKE with the WRF-ARW model using a horizontal resolution of 1 km and fairly high vertical resolution (~25 m in lowest 2 km, ~250 m in next 8 km, then ~500 m). The IC and<BR>
LBCs are provided by the NCEP NAM model. I am currently testing the TKE diffusion option (km_opt = 2) with<BR>
diffusion calculated in physical space (diff_opt=2) and the PBL scheme turned off (bl_pbl_physics = 0).<BR>
Unfortunately these runs are producing NANs in the output file for multiple variables. I have tested with both surface layer formulations (sf_sfclay_physics = 1 and 2), but I see no improvement. I have however had success when running with the Smagorinsky 2D deformation scheme (km_opt=4) and TKE provided by the MYJ PBL scheme (bl_pbl_physics=2). The time step is 3 and 1/3 seconds.<BR>
<BR>
My question: Can the diffusion TKE option be used at 1 km? Are there other settings that should be changed to use it?<BR>
<BR>
Thanks,<BR>
<BR>
Eric Kemp<BR>
Northrop Grumman Corporation/TASC<BR>
<BR>
</FONT>
</P>
</BODY>
</HTML>