1. SGP BORCAL Overview September 2005 Craig Webb, Ibrahim Reda, & Tom Stoffel U.S. DOE Atmospheric Radiation Measurement (ARM) Program

2. OUTLINE Introduction ARM Program Needs Broadband Outdoor Radiometer Calibrations Methodology Radiometer Calibration & Characterization Functionality Quality Assurance of the BORCAL Process Control Radiometer History Summary

3. Improve Global Climate Modeling

4. Advance Climate Change Research by Providing Accurate Measurements Most ambitious meteorological measurement program since the International Geophysical Year of 1957.

5. ARM Climate Research Facilities

6. Alaska Oklahoma Nauru ARM Mobile Facility Total of 29 Stations 250 + Radiometers

7. In-Situ Cloud Measurements ProteusAltus UAV “Digital” CM-22 & CG-4

8. Instantaneous Radiative Flux Cloud Parameterization and Modeling Cloud Properties Aerosol Properties Shortwave 3D Rad Tran Longwave / Water Single Column 2D/3D Models Cloud Properties Aerosol Working groupsSubgroups Radiometry

9. Radiometer Calibration Facility Lamont, Oklahoma USA

10. ARM Radiometer Calibration Facility 100 Pyranometers 30 Pyrheliometers In each of 2 BORCAL Events per Year

11. ARM Radiometers: Shortwave Calibration Traceability

12. ARM Radiometers: Longwave Calibration Traceability

13. Broadband Outdoor Radiometer CALibration Based on component summation… BORCAL is the Process

14. Component Summation Technique Direct Beam (Cavity Radiometer) * Cos(Z) = + Reference Diffuse (Shaded Pyranometer) Global Reference

15. But there is NO Reference for Diffuse! Eppley Model PSP & 8-48 Pyranometers

16. Shade / Unshade Calibration Standard Procedure: Rs= US - S Dir * Cos(Z) where, Rs = Pyranometer Responsivity (uV/Wm -2 ) US = Unshaded signal (uV) S = Shaded signal (uV) Dir = Direct Normal (Wm -2 ) Z = Solar Zenith angle Shade / Unshade cycles based on pyranometer time response

17. Time (mm:ss) 00:0035:00 Signal (mV) 0 10 Direct Unshade Shade Modified Shade / Unshade Calibration

18. RCC is the Software Mechanism Radiometer Calibration & Characterization Used to control the BORCAL process with Lab Windows and Access programs…

19. RCC Schematic

20. Responsivities 1.Rs (45-55) Traditional - Pyranometers Mid-latitude USA (following Ed Flowers & Don Nelson/NOAA) 2.Rs (Composite) Traditional - Pyrheliometers (results depend on amount of data collected) 3.Rs (9 deg intervals) Discrete Characterization - Improve information available to the user.

21. Responsivities 4. Rs (45 +/-0.3) Improved trend analyses by more repeatable results 5. Rs (2 deg interval) Improved Characterization 6.Rs (Z) Polynomial fit to AM & PM responses 7.Rs (Lat) Single Rs optimized for measurement station location

22. Event Configuration Configuration is accomplished through a hierarchical data base selection process for instruments and system parameters. Configuration is summarized in the top-level window Information is drawn from the underlying Access data tables

23. Configuration Verification Configuration verification is conducted by a two-person team. For each instrument, the configuration is verified using a prescribed process of visual inspection and data logging. Step 1: The instrument ID is visually verified and relayed to the computer operator. Verification Path: Visual System Configuration Data Logging Step 2: The instrument ID located in the RCC system and logging commenced. Step 4: The instrument location and cabling is visually verified and compared that in the RCC configuration. Step 3: The instrument is shaded for several seconds. The computer operator confirms the telltale drop in instrument voltage. Step 5: The instrument is marked as verified.

24. Data Acquisition Operator logging GPS or manual time maintenance Atmospheric stability monitor Real-time graphing of responsivity, irradiance, or voltage Real-time error trapping and diagnosis Detailed reference irradiance data Real-time access to data acquisition and quality control

25. Responsivity Calculations During the calibration event, responsivity is calculated at 30-second intervals as: RS = Thermopile Voltage / Reference Irradiance Direct Beam Reference Irradiance Measured by the absolute cavity radiometer. Diffuse Reference Irradiance Measured by two Eppley 8-48 pyranometers Global Reference Irradiance Calculated from Direct Beam and Diffuse Reference Glo = Dir Cos(zenith) + Dif Pyrheliometers Shaded Pyranometers* Pyranometers *Experimental Reference Irradiance

26. BORCAL Results Normal Incident Pyrheliometer (NIP)

27. BORCAL Results Pyranometer ± 4% ± 0.9% for 1º bin Rs (45º) 8.10 µV/Wm -2 ± 4%

28. Calibration Certificates Traceability and Certification Calibration ConditionsCalibration Results

29. Suggested Methods of Applying Results 45-55 Degree Responsivity Composite Responsivity Calibration History Latitude- Optimized Responsivity Responsivity Function 9-Degree Responsivities

30. Instrument Responsivity Pyranometer Responsivity Mean of all 30-second RS at 45° ±0.3° Responsivity (RS) is Calculated from the 30-second Individual Instrument Responsivities Shaded Pyranometer Responsivity* Mean of all 30-second RS Pyrheliometer Responsivity Mean of all 30-second RS at 45° ±0.3° *Experimental

31. Additional Pyranometer Characterizations 2-degree Bins Mean of all 30-second RS ±0.3° at 2° zenith angle increments Responsivity Function: RS(z) Polynomial in cos(z), fitted to all available 2-degree responsivities 45-55 Degree Mean of responsivities calculated from RS(z) between 45° - 55° 9-degree Bins Mean of responsivities calculated from RS(z) over 9° wide intervals Composite Cosine weighted from z = 0° to 90° of RS(z) Latitude Optimized Latitude limited, calculated from RS(z) and latitude

32. ~ 0.25% - 2.5% ~ 0.06% ~ 0.4% WRR Transfer of Direct Beam Irradiance Sources of Calibration Uncertainty Data Logger Zenith Angle Calculations (< 75°) Diffuse Sky Irradiance (w.r.t. reference global) Base Uncertainty for each data point as Root Sum Square of Sources of Uncertainty (with respect to reference irradiance) Pyrheliometers ~ 0.5% Pyranometers ~ 0.8% – 3.0% ~ 0.12% Excludes zenith angle and diffuse irradiance uncertainties Includes zenith angle and diffuse irradiance uncertainties

33. Pyranometer Calibration Uncertainty Terms U avg = Mean of base uncertainties (%) U std = Standard deviation, base uncertainties RS max = Highest responsivity (all data) RS min = Lowest responsivity (all data) RS= Mean responsivity @ 45° Intermediate Calculations U rad = [ U avg 2 + (2 U std ) 2 ] 1/2 E + = 100 (RS max – RS) / RS E – = 100 (RS – RS min ) / RS ±Uncertainties (%) U 95+ = +(U rad + E + ) U 95– = –(U rad + E – ) Calculated from 30-second data points

34. Pyrheliometer Calibration Uncertainty Terms U avg = Mean of base uncertainties (%) U std = Standard deviation, base uncertainties RS max = Highest responsivity (all data) RS min = Lowest responsivity (all data) RS= Mean responsivity (all data) Intermediate Calculations Range= 100 (RS max – RS min ) / RS Uncertainty (%) U 95 = [ U avg 2 + (2 U std ) 2 + (Range/2) 2 ] 1/2 Calculated from 30-second data points

35. Database Maintenance Tools Data Editing Forms Data Access Data Plotting Tools Calibration Results System Data Tables Instrument Inventory Customer Calibration Results System Configuration Calibration Facility Data Acquisition

36. Data Exporting and Distribution Responsivity Data Export for Transferring Calibration Results Custom Exports with Selectable Parameters and Output Format Dedicated Export Format (AIM Database, Calibration Stickers)

37. ARM Instrument Management (AIM) System

38. Confirming the Process 11-year History of Control NIP Calibrations BORCAL Event1995 2005 8.70 8.40 Rs (uV/Wm -2 ) 2.0%

39. Confirming the Process 11-year History: Control PSP Calibrations (Set 1) BORCAL Event1995 2005 9.10 7.60 Rs (uV/Wm -2 ) 5.0%

40. Confirming the Process 11-year History: Control PSP Calibrations (Set 2) BORCAL Event1995 2005 8.90 7.80 Rs (uV/Wm -2 ) 5.6%

41. Summary The Atmospheric Radiation Measurement (ARM) Program needs accurate broadband irradiance data from three climatic regions. NREL has developed RCC software to semi- automate the BORCAL process for hundreds of shortwave radiometer calibrations each year. The AIM Database contains BORCAL and deployment information for all ARM radiometers (http://www.nrel.gov/aim) Control Radiometer calibration history shows long- term performance of BORCAL process and radiometer responsivities.

42. Thanks! Questions for Craig or Reda? More information available from http://www.arm.gov