SCMS95: BARO Messages: 2 Entries..

BARO_IC REPORT: the format for this output is contained in file: baro_ic.r



 8/13/95              HARP Barometer Intercomparisons                 Pg   1


Sta      Date    IC #   Baro #      IC     PAM      Diff      Comments
---   ----------------  ------    ------  ------    ----      --------
  1    7/19/95   37039     3      1012.0  1012.3     -.3      About .3  low
       7/28/95   37039     3      1016.5  1016.6     -.1      good
       8/07/95   37039     3      1018.0  1018.3     -.3
       8/13/95   37039     3      1015.4  1015.6     -.2

  2    7/20/95   37039     2      1016.0  1016.3     -.3      .3 mbar off.  I
       7/28/95   37039     2      1016.8  1017.0     -.3      .3mb high as us
       8/13/95             2      1016.2  1016.5     -.3

  3    7/11/95   37039     1      1016.1  1016.3     -.2      close
       8/06/95   paro      1      1017.5  1017.9     -.3
       8/13/95             1      1015.7  1016.0     -.3














































Post SCMS95 Pressure Sensor Report

Introduction:

	This report looks at the performance of the Vaisala pressure sensors
used in PAM III during SCMS95. Pre and post checks of the sensors are
looked at along with intercomparison data taken during the project.

Pre-Project tests:

	The pre-project testing consisted of running the sensors over a
temperature range of -15 to 35 C while varying the pressure from 70.0 to
110.0 kPa.  Table 1 below is the average and standard deviation of the
errors. The Ruska quartz pressure sensor was used as the reference.


sensor		mean error		standard dev.
		  (Pa)			   (Pa)

  1		  -4.6			   3.1
  2		  -1.8			   3.2
  3		  -0.5			   2.3


Post-Project tests:

	The sensors were checked against the dead weight gauge over the
range of 60.0 to 110.0kPa. The results are in table 2. Recent tests
of the dead weight shows an accuracy of about 5 Pa. Refer to dead weight
intercomparison paper for more information.

sensor		mean error		standard dev.
		  (Pa)			   (Pa)

  1		  -4.4			   2.6
  2		  ----			   ---
  3		  -0.4			   3.0

NOTE:  Sensor #2 had failed. A power supply lan on the circuit board had
burned up. The cause of this event is unknown. It appeared to happen
sometime between the end of SCMS95 and the post-project tests.


Intercomparison tests:

	During SCMS95 intercomparison tests were run using the
ParoScientific as a reference.  It became obvious during these tests
that a systematic offset existed between the Paro and the PAM III sensors.
A test of the Paro against the dead weight showed a bias of 30.79 Pa
(.3079 mb) low. Table 2 below shows the intercomparison data with
and without the offset.

station		sensor		error w/ offset		error w/o offset
				    (Pa)		     (Pa)
  1		  3		   -23.0		    8.0
  2		  2		   -30.0		    1.0
  3		  1		   -27.0		    4.0

Conclusion:

	The Vaisala pressure sensor performed as we expected.
The reason for the failure of sensor 2 is unknown.  A post check
will be done once the sensor has been repaired.


NOTE: all error terms are defined as reference - sensor.