DQRID : D060824.1
Start DateStart TimeEnd DateEnd Time Data Quality Metric
08/11/2006193008/25/20061859Incorrect
more
Subject:
NSA/GNDRAD/C1 - Incorrect calibration coefficients
DataStreams:nsagndradC1.00, nsagndrad20sC1.a1, nsagndrad60sC1.b1, nsagndrad20sC1.a0
Description:
BarrowC1, Gndrad. On Aug 11th the PSP-US was changed out and the Campbell program was 
changed to reflect this.  However, an old program that uses the 4 coefficient PIR cal factors 
was used. The PIR-UIR was not changed out, however, the logger program coeffs reverted 
back to the 4 Coeffcients.  The Eppley PIR-UIR coefficients for this S/N 30168F3 are :  
K0=0.0, K1(aka k)=.25253, K2= 1.0, K3=-4, Kr=0.0.
Suggestions: 
Either disregard the Gndrad PIR data for this time period or recalculate 

up_long_hemispheric irradiance based on the following information:

For a PIR mounted in an upwelling measurement orientation (UIR), the
thermopile output is extremely low compared to a downwelling PIR (DIR).
In addition, the UIR case and dome temperature differences are much
smaller than a DIR.  This results in the pyrgeometer receiver radiation 
term, Wr (NREL/Hickey equation) or Wc (Albrecht/Cox equation), dominating
all other terms of the infrared radiant flux equation.

Since the receiver radiation term uses coefficients in both forms of
the pyrgeometer equations nearly equal to 1.0 +- <1% the result is a
measurement of incoming longwave radiant flux relatively insensitive 
to the installed instrument - so long as it is an Eppley PIR.
		
Evaluation of the problem data shows the expected error of the data
(before correction) will be within 1% - less than the uncertainty of 
the instrument (Eppley PIR greater of 5% or 5 W/m2)
		
To correct the one minute data found in the a1 or b1 datastream follow
this procedure:
1) Obtain the following variables from the b1 datafile:
	 inst_up_long_dome_temp
	 inst_up_long_case_temp
	 up_long_netir

2) For each one minute data entry in b1 backout the original PIR thermopile reading in uV using the incorrectly applied K1 coefficient.
     tpUIR[uV] = up_long_netir[uV]/0.2582[W/m2/uV]

3) Apply for each one minute data entry:

WinUIR(W/m2)=k*V+ Sigma*Tc^4 -4*Sigma*(Td^4 - Tc^4)
	    =[0.25253*tpUIR]+[Sigma*inst_up_long_case_temp^4]
	      -[4*Sigma*(inst_up_long_dome_temp^4-inst_up_long_case^4)]

where, Sigma =	5.6697E-8 (Stefan-Boltzman Const)
		
Other datastream corrections (e.g. to a0) that use the UIR measurements will require similar corrections.
      
For additional information see:
http://www.arm.gov/publications/tech_reports/arm-05-111.pdf

























Either disregard the Gndrad PIR data for this time period or recalculate 
up_long_hem
ispheric irradiance based on the following information:

For a PIR mounted in an upwelling measurement orientation (UIR), the
thermopile output is extremely low compared to a downwelling PIR (DIR).
In addition, the UIR case and dome temperature differences are much
smaller than a DIR.  This results in the pyrgeometer receiver radiation 
term, Wr (NREL/Hickey equation) or Wc (Albrecht/Cox equation), dominating
all other terms of the infrared radiant flux equation.

Since the receiver radiation term uses coefficients in both forms of
the pyrgeometer equations nearly equal to 1.0 +- <1% the result is a
measurement of incoming longwave radiant flux relatively insensitive 
to the installed instrument - so long as it is an Eppley PIR.
		
Evaluation of the problem data shows the expected error of the data
(before correction) will be within 1% - less than the uncertainty of 
the instrument (Eppley PIR greater of 5% or 5 W/m2)
		
To correct the one minute data found in the a1 or b1 datastream follow
this procedure:
1) Obtain the following variables from the b1 datafile:
	 inst_up_long_dome_temp
	 inst_up_long_case_temp
	 up_long_netir

2) For each one minute data entry in b1 backout the original PIR thermopile reading in uV using the incorrectly applied K1 coefficient.
     tpUIR[uV] = up_long_netir[uV]/0.2582[W/m2/uV]

3) Apply for each one minute data entry:

WinUIR(W/m2)=k*V+ Sigma*Tc^4 -4*Sigma*(Td^4 - Tc^4)
	    =[0.25253*tpUIR]+[Sigma*inst_up_long_case_temp^4]
	      -[4*Sigma*(inst_up_long_dome_temp^4-inst_up_long_case^4)]

where, Sigma =	5.6697E-8 (Stefan-Boltzman Const)
		
Other datastream corrections (e.g. to a0) that use the UIR measurements will require similar corrections.
      
For additional information see:
http://www.arm.gov/publications/tech_reports/arm-05-111.pdf











Either disregard the Gndrad PIR data for this time period or recalculate 
up_long_hemi
spheric irradiance based on the following information:

For a PIR mounted in an upwelling measurement orientation (UIR), the
thermopile output is extremely low compared to a downwelling PIR (DIR).
In addition, the UIR case and dome temperature differences are much
smaller than a DIR.  This results in the pyrgeometer receiver radiation 
term, Wr (NREL/Hickey equation) or Wc (Albrecht/Cox equation), dominating
all other terms of the infrared radiant flux equation.

Since the receiver radiation term uses coefficients in both forms of
the pyrgeometer equations nearly equal to 1.0 +- <1% the result is a
measurement of incoming longwave radiant flux relatively insensitive 
to the installed instrument - so long as it is an Eppley PIR.
		
Evaluation of the problem data shows the expected error of the data
(before correction) will be within 1% - less than the uncertainty of 
the instrument (Eppley PIR greater of 5% or 5 W/m2)
		
To correct the one minute data found in the a1 or b1 datastream follow
this procedure:
1) Obtain the following variables from the b1 datafile:
	 inst_up_long_dome_temp
	 inst_up_long_case_temp
	 up_long_netir

2) For each one minute data entry in b1 backout the original PIR thermopile reading in uV using the incorrectly applied K1 coefficient.
     tpUIR[uV] = up_long_netir[uV]/0.2582[W/m2/uV]

3) Apply for each one minute data entry:

WinUIR(W/m2)=k*V+ Sigma*Tc^4 -4*Sigma*(Td^4 - Tc^4)
	    =[0.25253*tpUIR]+[Sigma*inst_up_long_case_temp^4]
	      -[4*Sigma*(inst_up_long_dome_temp^4-inst_up_long_case^4)]

where, Sigma =	5.6697E-8 (Stefan-Boltzman Const)
		
Other datastream corrections (e.g. to a0) that use the UIR measurements will require similar corrections.
      
For additional information see:
http://www.arm.gov/publications/tech_reports/arm-05-111.pdf
Measurements:nsagndrad60sC1.b1:
  • up_long_hemisp_min
  • net_std
  • pir1_min
  • net_sd
  • up_long_hemisp
  • net_mean
  • pir1_mean
  • net_min
  • up_long_hemisp_std
  • up_long_hemisp_max
  • pir1_max
  • pir1_sd
  • net_max
more
nsagndradC1.00:
  • Raw data stream - documentation not supported
more
nsagndrad20sC1.a1:
  • pir1_voltage
  • pir1_dome_therm
  • inst_up_long_case_resist
  • inst_up_long_hemisp_tp
  • pir1_case_therm
  • inst_up_long_dome_resist
more
nsagndrad20sC1.a0:
  • inst_up_long_case_resist
  • inst_up_long_dome_resist
  • inst_up_long_hemisp_tp
more

Close this window