DNMI INTERCOMPARISON REPORT FOR BAROMETERS

Dr. Theo Schaad 

24. September 1998   
  Summary

A five-year old Digiquartz® Barometric Standard was compared with two primary pressure standards at the Paroscientific Calibration Laboratory as part of an international calibration intercomparison originated by the Norwegian Meteorological Institute DNMI. The five-year stability of the device was –1.0 Pa per year (-0.010 hPa/year). The conformance (linearity) “as found” from 600 to 1100 hPa was ±1.0 Pa (DH Standard) and ±1.2 Pa (Bell & Howell Standard). Both results are 2-sigma values including measurement errors. 

Paroscientific, Inc., is the world leader in barometric instrumentation. The Digiquartz® barometric standards and the Digiquartz® MET3A meteorological measurement system provide the ultimate in precision meteorological measurements. The intercomparison results attest to the highest degree of measurement reliability at the Paroscientific Calibration Laboratory.  They show experimentally that actual performance of this transfer standard far exceeded specifications at levels usually reserved for micro barometry. The Digiquartz® Barometric Standard is the ideal high-precision portable instrument for calibration intercomparison both at the laboratory and in field use.

  Introduction

The Norwegian Meteorological Institute DNMI arranged a barometric intercomparison on behalf of the Nordic Conference for Measurement and Calibration. The intercomparison is based on a portable barometric transfer standard manufactured by Paroscientific, Inc. The chosen instrument has a well-documented calibration history extending over five years, which allows for a critical comparison of measurement techniques and systematic differences between laboratories. DNMI asked the participating laboratories to perform two tests at ambient laboratory temperature. The first one is an assessment of measurement scatter by repeating a particular pressure sequence 10 times. The second test is a conformance (linearity) measurement of at least 10 pressure values in ascending and descending direction (2 times) from 600 to 1100 hPa. The data is taken by “normal” calibration practices and not optimized to reduce measurement error. 

  Testing Standards

Transfer Standard:    Digiquartz® Barometric Standard

                                    Model 1016B-01

                                    (Special range 600 to 1100 hPa)

Calibration Date:          19 January 1993

Offset Adjustment:        0.057 hPa (9 September 1998)

Accuracy:                     Better than 0.01 % of reading

(including conformance, repeatability, and hysteresis)

Product Information:     See http://www.paroscientific.com

Pressure Standard:    Bell & Howell Absolute Dead Weight Tester 100 to 3400 hPa (absolute)

Calibration Date:          11 December 1996

Accuracy:                     0.01 % of reading, traceable to NIST

Note:                            Original equipment used to calibrate transfer standard in 1993

 

Pressure Standard:    DH Instruments PG7601 Piston Gauge

                                    50 to 3500 hPa (absolute)

Calibration Date:           29 May 1997 (purchase date)

Accuracy:                     ± (20 ppm + 0.2 Pa) (piston)

                        ± 10 ppm (weights)

                                    traceable to NIST                                

The transfer barometer was placed in a passive thermally protected chamber at ambient laboratory conditions. The test device was continuously powered by an 8.0 volt AC/DC adapter, starting one hour prior to the first measurement. The applied pressure was corrected for actual mass values, vacuum back-pressure, piston temperature, local gravity, and height between the test device and the reference. The weight bounce of the dead weight testers was monitored with the high-resolution output of the transfer barometer. Data was only taken after the vacuum dropped below 2.7 Pa (Bell & Howell) or 4 Pa (DH Piston Gauge). Typical setup times were 6 minutes per measurement. Only single measurements were taken without averaging.
Pressure Units   
The measurements are reported in metric pressure units. The fundamental unit is the Pascal (Pa). The commonly used meteorological unit is hectoPascal (1 hPa=100 Pa), which is identical to millibar (1 mbar=1 hPa).  

hPa

mbar  

Pa

microbar  

percent of

1000 hPa

parts per million

in Hg  

0.1 10 100 0.01% 100 ppm 0.00295
0.01 1 10 0.001% 10 ppm 0.00030
0.001 0.1 1 0.0001% 1 ppm 0.00003
Test Results   

Stability: The results are strictly reported “as found”, except for a single offset adjustment of 0.057 hPa, which was established with the same working standard that was used in the original calibration in January 1993. The averaged stability was –0.01 hPa per year, which is by itself remarkable. 

Repeatability: The measurement repeatability includes the contributions of the intrinsic repeatability of transfer standard and pressure standards, the stability of the applied pressure, and the resolution of the measurement. The measurement repeatability was 0.3 Pa (standard deviation) for both pressure standards. 

Hysteresis: Pressure hysteresis was measured by approaching mid-scale from low and high limits of the pressure range. The measured value was 0.0 ± 0.2 Pa, consistent with measurement scatter (no measurable hysteresis). 

Conformance: The conformance “as found” is shown graphically in the attached graphs. There is no evidence for non-linearity. The 2-sigma limit of the error band is:  

Pressure Standard  

Conformance  

Bell & Howell Tester  

< 0.012 hPa  

DH PG7601 Piston Gauge  

< 0.010 hPa  

Bell&Howel

PG7601

©2007 Paroscientific, Inc.