[an error occurred while processing this directive]

  <p><img style="float:left" src="/images/research/themes.jpg" alt="themes" /></p>

  <p>&nbsp;<br />&nbsp;<br />&nbsp;<br />&nbsp;<br />&nbsp;<br /></p>

  <table width="60%" border="0" cellspacing="0" cellpadding="0">
  <tr><td>

  <h2>Physical Processes Group</h2>
  <span style="color:#337799"><b>From</b> <a href="mailto:drosa@berkeley.edu">
  Daniele Rosa</a></span><br />
  <span style="color:#337799"><b>Date</b> June 21, 2012</span><p></p>

Hi<p></p>

reading the SPCAM code it seems that local and non-local diffusion is computed by the standard CAM 3.5 PBL parameterization and then the CRM is allowed to do more mixing on constituents in the PBL layers.<p></p>

So SPCAM does not replace the conventional PBL mixing, instead it adds some.<p></p>

Do I get it right?<p></p>

Thanks<br />
Daniele<br />

  <br /><hr /><p></p>

  <span style="color:#337799"><b>From</b> <a href="mailto:mspritch@u.washington.edu">
  Mike Pritchard</a></span><br />
  <span style="color:#337799"><b>Date</b> June 21, 2012</span><p></p>

Hi Daniele,<p></p>

Awesome question.<p></p>

My understanding from closing the T,Q,U & V budgets in the SPCAM3.0 code was that the PBL routine is bypassed for Q & T diffusion but is used to do U & V diffusion (at least in the standard configuration of a 2D CRM; if the CRM is 3D the PBL scheme is entirely bypassed and the resolved momentum transport from the 3D CRM is used instead). I can't fully recall if this was also the case for SPCAM3.5, but I thought so. <p></p>

Actually, I think this might be an important lingering issue for some of the many SPCAM runs that have been run on greater than 26 vertical levels - Rich Neale has mentioned in the past that the CAM3.5.X PBL scheme has bugs that make it behave unphysically on anything other than the L26 vertical grid. So it's quite possible that SPCAM3.5 simulations that have used grids other than L26 may have suffered from unphysical diffusive momentum transport. <p></p>

Cheers,<p></p>

-Mike. 

  <br /><hr /><p></p>

  <span style="color:#337799"><b>From</b> <a href="mailto:marat.khairoutdinov@stonybrook.edu">
  Marat Khairoutdinov</a></span><br />
  <span style="color:#337799"><b>Date</b> June 21, 2012</span><p></p>

Daniel,<p></p>

The vertical countergradient PBL transport is switched off in src/physics/cam/diffusion_solver.F90 when CRM is on. Some CAM grid diffusion is left to maintain CAM's advection stability and for physical reasons (as for stable stratification, SAM is not that good in mixing and diffusing things around).<p></p>

Marat

  <br /><hr /><p></p>

  <span style="color:#337799"><b>From</b> <a href="mailto:drosa@berkeley.edu">
  Daniele Rosa</a></span><br />
  <span style="color:#337799"><b>Date</b> June 22, 2012</span><p></p>

Marat:<p></p>

How is it done in the code? I can't find any "#ifdef CRM" or anything else.<p></p>

Is there an equivalent PBL diffusion in SAM when it is run by itself?<p></p>

Daniele

  <br /><hr /><p></p>

  <span style="color:#337799"><b>From</b> <a href="mailto:marat.khairoutdinov@stonybrook.edu">
  Marat Khairoutdinov</a></span><br />
  <span style="color:#337799"><b>Date</b> June 22, 2012</span><p></p>

try to look for #ifndef CRM, or just CRM. -Marat<p></p>


<br /><hr />


  </td></tr></table>

[an error occurred while processing this directive]
</body>
</html>