| MTP Home Page  | 

Tropopause Intercomparison on SONEX



Six SONEX flights were chosen to perform a comparison of tropopause altitudes based on radiosondes (RAOBs), potential vorticity (PV), LaRC/DIAL ozone mixing ratio, and JPL/MTP temperature. In the table below, the first and last flights (19971013 and 1997115) -- the two transit flights across the US -- were used to calibrate the MTP against radiosondes and determine the PV tropopause threshold (3.0).

Clicking on the flight date will link to a GIF image containing the following information:
1) A green trace representing the DC8 pressure altitude (pALT)
2) A light blue trace representing the PV tropopause
3) A dark blue trace representing the MTP tropopause (and a red trace if a second tropopause is present)
4) A yellow trace representing the DIAL ozone gradient tropopause
5) A cream trace (at 1 km) representing the DIAL tropopause fold flag
6) A green trace offset from 2 km representing the difference between pressure and geometric altitude
7) Red squares representing the RAOB tropopause measurements. In general RAOBs at the beginning or end of the flights are the preceeding or following sondes, while those during a flight are at the proper time. In these plots, no attempt has yet been made to show the spatial location of the RAOB launch site with respect to the DC8 location. As a result some RAOBs disagree significantly with the other tropopause determinations. For example, the RAOB in the middle of the 1991110 flight is for Inukjuak, which was well north of the DC8's flight track to Hudson's Bay, and had a much lower tropopause.

Some general conclusions:

1) In general MTP and RAOBs agree quite well.
2) Ozone shows much more variability in tropopause height compared to other methods (see for example, 19971018).
3) In regions of frontogenesis PV and Ozone change much faster than MTP (see for example, 19971020).
4) There does not appear to be a trend for either PV-based or ozone-based tropopause determinations to be systematically high or low with respect to MTP.

Suggestions?



 
19971013
19971018
19971020
19971023
19971028
19971109
19971110
19971115