Stratosphere

The FV3 model uses a hybrid sigma vertical coordinate.  Near ground, the model vertical coordinates
are sigma near the surface, pressure levels in the stratosphere and a blend in between. See
ak_bk.txt for the exact values in determining the vertical levels.

The PRES is the pressure which includes hydrostatic and dynamic elements.  These are the mid-layer
values of PRES < 200 hPa for a typical time.

2024030818:PRES:51-52 hybrid pressure layer:15 hour fcst::mean=1944.84:min=1943.55:max=1946.92:cos_wt_mean=1944.87
2024030818:PRES:52-53 hybrid pressure layer:15 hour fcst::mean=1646.41:min=1644.46:max=1647.57:cos_wt_mean=1646.44
2024030818:PRES:53-54 hybrid pressure layer:15 hour fcst::mean=1384.6:min=1383.26:max=1386.37:cos_wt_mean=1384.63
2024030818:PRES:54-55 hybrid pressure layer:15 hour fcst::mean=1155.08:min=1154.25:max=1156.63:cos_wt_mean=1155.1
2024030818:PRES:55-56 hybrid pressure layer:15 hour fcst::mean=953.968:min=953.392:max=954.659:cos_wt_mean=953.99
2024030818:PRES:56-57 hybrid pressure layer:15 hour fcst::mean=777.848:min=777.452:max=778.478:cos_wt_mean=777.867
2024030818:PRES:57-58 hybrid pressure layer:15 hour fcst::mean=623.656:min=623.193:max=624.195:cos_wt_mean=623.672
2024030818:PRES:58-59 hybrid pressure layer:15 hour fcst::mean=488.719:min=488.228:max=489.541:cos_wt_mean=488.733
2024030818:PRES:59-60 hybrid pressure layer:15 hour fcst::mean=370.611:min=369.729:max=371.318:cos_wt_mean=370.621
2024030818:PRES:60-61 hybrid pressure layer:15 hour fcst::mean=267.218:min=266.112:max=268.199:cos_wt_mean=267.23
2024030818:PRES:61-62 hybrid pressure layer:15 hour fcst::mean=176.538:min=175.802:max=177.576:cos_wt_mean=176.542
2024030818:PRES:62-63 hybrid pressure layer:15 hour fcst::mean=96.3802:min=95.9721:max=96.8012:cos_wt_mean=96.3874
2024030818:PRES:63-64 hybrid pressure layer:15 hour fcst::mean=37.9142:min=37.8266:max=38.0601:cos_wt_mean=37.9148

Modern radiosonde can reach 3 hPa.  Given the smooth nature of straospheric analyses, you typically don't need
a high density of observations.

Sondes usually need a stratospheric temperature bias correction for solar heating.  The correction is either
be done at the station or by data assimilation system.  The bias correction depends on the instrument, and
given the age of these observations, the bias correction may not be as accurate as you want.



GPS-RO: high accuracy meaurements using signals from global positioning satellites.  We had a debate whether to use 
GPS-RO.  GPS-RO uses time information, so it high accuracy.  The disadvantage of using GPS-RO is it is available only 
from late 1990s onwards, the coverage is much different from the sonde network, and there is the potential to introduce 
a "climate shift" by introducing a large number of observations in a region that the relatively poorly observed.
Ideally one would run a version of CORe starting from 199x assimilating the GPS-RO data to get an improved stratosphere.
Of course, one would want the newer version of FV3GFS with twice the number of vertical layers.