het_chem

het chem project update(s)

To do:

• Compare MIPAS and WACCM profiles
• 70-30 hPa
• Plot individual years
• Apply changes to monsoon
• Spatial deNO_y

1. NH Subpolar
   • Profile discrepancy
   • UTLS (164-68 hPa)
     • MIPAS HNO₃ PDFs
     • MIPAS HNO₃ CDFs
   • 70-30 hPa
     • MIPAS HNO₃ PDFs
     • MIPAS HNO₃ CDFs
   • Individual years
2. Monsoon HNO₃

NH Subpolar

Profile discrepancy

There are substantially more profiles in WACCM compared to MIPAS. However, good days look, well, good:

Red circles are WACCM, black circles are MIPAS. Pretty good match. But the problem is all the downtime that MIPAS had over its lifetime. The plot below shows the number of NH profiles for each day in WACCM (red), MIPAS (black), and the daily difference (blue):

Essentially, there are many days where MIPAS is observing nothing but WACCM is observing the usual ~500 profiles in the NH. I changed my scripts to only report WACCM profiles on days when MIPAS also has at least 1 profile inside the vortex edge. As you'll see below, this helps but there are still more WACCM profiles than observed. So I may have to set a higher threshold (in the map above there are ~50 profiles...maybe 15 or more is a "good" day? I can look for a "bad" MIPAS day and plot that vs. WACCM), or I can match exact profile numbers.

UTLS (164-68 hPa)

UTLS MIPAS HNO₃ profiles and PDFs

Here are Jan, Feb, and March respectively:

This does not explain the big peak in hetAll.008, but I'm thinking to take, e.g., the first 15 days of March only (similar to how we were doing 0901-1015 for SH at one point).

UTLS HNO₃ CDFs

In all cases, all the CDFs are drawn from different distributions (and it looks like indeed noHetAll is closest to the obs):

70-30 hPa

MIPAS HNO₃ profiles and PDFs

Same as above, but the distribution plots are for 70-30 hPa. After looking at this, I could probably go to ~20 hPa. Anyway, these look really good! You were right Doug, looks like this is where the magic happens.

HNO₃ CDFs

Individual years

Below are the indiviual years for February. In general, WACCM varies a lot more than the obs.

Monsoon

Here are the PDF plots for June-August for 0-30N, Eastern Hemisphere only:

Here are Jan, Feb, and March respectively:

And the CDFs:

Monsoon looks good, I think, and HetAll.008 certainly looks the best out of the WACCM bunch. It looks like this could be a bias issue, and I'll try shifting means tomorrow and see how the CDFs look. July 2011 is also shown below as an example of a single year: