The SKYNET Radiometer Network: Aerosol Optical Depth Retrieval

The SKYNET radiometer Network: Aerosol Optical Depth retrieval performance at the FRC‐IV campaign and long‐term comparison against GAW‐PFR and AERONET standard instruments

M. Campanelli1, V. Estellés2, H. Diemoz3, N. Kouremeti4, S. Kazadzis4, R. Becker5, S. Vergari6, S. Dietrich1

1 ISAC‐CNR, Rome Italy [email protected] 2 University of Valencia, Spain 3 Arpa Valle D’aosta, Italy 4 PMOD/WRC, Davos, Switzerland 5 DWD, Lindemberg, Germany 6 Italian Air Force, Rome, Italy

1. Introduction

Among the objectives of the World Meteorological Organization (WMO), as established by the Commission for Instruments and Methods of Observation (CIMO) is: “ensuring sustained high‐ quality meteorological observations […] in a changing world”. Assessing the quality and traceability of data sourced from WMO partners and private networks, and performing regular intercomparisons for characterizing and testing instruments and methods of observation, are therefore key issues for obtaining an accurate meteorological database, that is essential to assess the climate change. Quality and traceability is required by WMO also for measurements of atmospheric aerosols optical properties, performed by International networks of radiometers, because aerosol plays a key role in earth radiative balance and its response to climate change may constitute a relevant feedback loop (IPCC 2013). SKYNET (Takamura et al, 2004) is an International network with headquarter located in Japan, dedicated to studies on aerosols and their interaction with clouds and solar radiation. It consists of about 60 sites and it is organized in several Regional subnetworks. The European Skynet Radiometers network (ESR, www.euroskyrad.net, Campanelli et al, 2012) is the European Regional subnetwork, born in 2010 and managed by ISAC/CNR (Institute of Atmospheric Sciences and Climate, National Research Council, Italy) and the University of Valencia (Spain). SKYNET has been recently included as WMO –GAW (Global Atmospheric Watch) contributing network and a program of traceability to CIMO defined standard instruments and elaborations algorithms, together with intercomparisons and calibration of masters instruments, is important in order to make the SKYNET products homogeneous and comparable with other WMO‐GAW contributing networks. Among them there is the GAW‐PFR network, which is directly traceable to the PMOD/WORCC (Physikalisch‐ Meteorologisches Observatorium Davos/ World Optical depth Research and Calibration Center). In this regard, two ESR/SKYNET sun‐sky radiometers took part to the Filter Radiometer Comparison (FRC) held in Davos, Switzerland from September, 28 to October, 16, 2015. Furthermore, a comparison from three co‐located different radiometers of ESR/SKYNET, POM01, AERONET and GAW‐PFR networks was performed at Burjassot (Valencia, Spain) from January to November 2015. In the present study the results of the inter‐comparison campaigns will be shown and a future plan of traceability study will be discussed.

2. Equipment

ESR is the European Regional Network of SKYNET and currently consists of 20 sites located in Europe (Italy, Spain, Germany, Poland, UK), USA, and Antarctica (Fig. 1). The standard instrument is the sun‐sky radiometer PREDE/POM, a scanning spectral radiometer with a field of view of 1 degree, performing measurements of solar direct and diffuse sky irradiance at several scattering angles and selected wavelengths (Table 1). The detector unit is held at a constant temperature of 30°C. More information on the network is available on the web page www.euroskyrad.net. Columnar estimation of aerosol optical depth (AOD) at all the available wavelengths, Angström exponent, aerosol size distribution, single scattering albedo, phase function, and complex refractive index are the products provided for each site in near real time, and downloadable from the web site. Instruments are monthly calibrated by the improved Insitu Langley Plot technique (ILP) (Campanelli et al. 2007, 2004) directly at the site where they are located. This methodology assures a more frequent check of the status of the equipment, reducing the expenses necessary for its frequently shipping to calibration centers or high mountain sites, which are considered to be the more appropriate calibration locations. The GAW‐PFR network (Wehrli, 2005) consists of 12 sites where Precision Filter Radiometers (PFR) are installed for the measurements of AOD. This network is the WMO reference for traceability of AOD measurements. The PFR is a classic Sun Photometer with 4 independent channels, a field of view of 2.5° and equipped with 3 to 5nm bandwidth interference filters. The detector unit is held at a constant temperature of 20°C by an active Peltier system. AERONET (Holben et al., 1998) is a federation of ground‐based remote sensing aerosol networks established by NASA and PHOTONS and is greatly expanded by networks and collaborators from national agencies, institutes, universities, individual scientists, and partners. The program provides a long‐term, continuous and readily accessible public domain database of aerosol optical, microphysical and radiative properties for aerosol research and characterization, validation of satellite retrievals, and synergism with other databases. The network imposes standardization of instruments, calibration, processing and distribution. The standard instrument is a CIMEL sun‐sky radiometer with functionality similar to the PREDE/POM with the exception of a different field of view (1.2 degree), the missing of the temperature control unit for the optics, and the use of two different sensors for measuring direct and diffuse solar irradiance.

Figure 1. Map of ESR network

Model POM01 Model POM02 Wavelengths (nm) 315, 400, 500, 675, 870, 315, 340,380, 400, 500, 675, 870, Filter band: 940, 1020 940, 1020, 1627, 2200 2nm(UV),10nm(VIS),40nm(NIR)

Table 1. Wavelengths for POM’s models

3. Methodology a) The FRC‐IV campaign

Two PREDE/POM‐02 of the ESR/SKYNET network took place to the IV Filter Radiometer Comparison (FRC) as part of the XII WMO International Pyrheliometer Comparison (IPC), held in Davos (at 1590 m above sea level) from 28 September to 16 October 2015. Both the instruments, one from Italy ( managed by ARPA Valle d’Aosta) and one from Germany (managed by DWD), were set to take measurements of sun direct solar radiation every 1 minute and of diffuse solar irradiance every 10 minutes. According to the suggestions of the FRC Organization, the intercomparison against the triad of Precision Filter Radiometers (PFR) installed at Davos should be performed between Aerosol Optical Depth (AOD) at the following wavelengths: 368±3, 412±3, 500±3, 865±5 nm. Therefore the wavelengths in common with POM02 are only 500 and 870 nm. Two different values of solar calibration constants for POM02 were used for the intercomparison: a) a pre‐campaign calibration obtained with the insitu ILP technique in the permanent site (Campanelli et al., 2007, 2004); b) a calibration obtained with the ILP technique during the campaign at Davos. Results from the two tests will be discussed in the following section. b) The long‐term comparison against GAW‐PFR and AERONET standard instruments

A PREDE/POM01 of ESR/SKYNET, a CIMEL/CE318 of AERONET (both managed by the University of Valencia), and a PFR (of PMOD/WORCC) were simultaneously operating in Burjassot site, a suburban site in the metropolitan area of Valencia (Spain), affected by both urban and maritime aerosols and by frequent dust intrusions of Saharan dust and occasional wild fires particles in summer. The PFR was deployed at Burjassot site during the period February – December 2015, although only the period February – April 2015 is considered for the comparison in order to match the three instrument databases. The calibration constant of PFR before February was applied in this study, the PREDE/POM01 was calibrated monthly with the Improved Langley method, and L2 level of AODs from CIMEL were considered. The wavelengths in common among the 3 instruments are 500 and 870 nm, although the comparison has been extended at other wavelengths (440, 675 and 1020 nm) by using the Anström exponent (AE) determined with the instrument wavelengths.

4. Results a) The FRC‐IV campaign

During the campaign there were five days (28, 29, 30 of September, 1 and 12 of October) with sunshine and only very limited presence of clouds. During the five days the AOD at 500 nm varied from 0.02 to 0.12, which can be considered as normal values for the area. As described in section 3, two calibration values were used for calculating AOD of the POMs: the pre‐campaign ILP ( only for the Italian instrument) and the Davos ILP ( for both the instruments). In spite of the low values of AOD (a negative aspect for this in situ calibration methodology), the ILP technique produced very good results, and showed a small variation of calibrations values before and after the shipment of the instruments (0.2% at 500 nm and 1.2%. at 870 nm, for the Italian POM02). This could be interpreted as a symptom that the instruments are affected by the transport, by producing small changes in the alignment or of the internal parts of the optic, bringing to the necessity of a new recalibration once the instrument is located for measurements. The result of the comparison against the PFR, as expected, depends on the calibration accuracy. Figure 2 shows the improvement of the comparison for the Italian POM after the re‐calibration performed in Davos for 500 and 870 nm. Figure 3 shows the comparison for the German POM. Results showed that the discrepancy between POMs and PFR is always within the thresholds established by WMO. A deeper analysis also showed that the algorithm used by ESR could be improved by introducing minor changes in the cloud screening procedure, mainly in the creation of the direct sun triplets, especially for wavelength 500 nm.

Figure 2: AOD differences between the Italian POM and PFR, using the calibration constants calculated in the previous site of measurements ( grey dots) and in Davos ( green dots). Grey curves are the thresholds of agreement established by WMO.

Figure 3: AOD differences between the German POM and PFR, using the calibration constants calculated in Davos (blue dots). Grey curves are the thresholds of agreement established by WMO. b) The long‐term comparison against GAW‐PFR and AERONET standard instruments

The average AOD and AE (based on AERONET data level 2 products) during the period of intercomparison, was 0.083 and 1.20, respectively, although occasional periods with AOD greater that 0.3 at 500 nm were recorded. The general conditions have been good for the validation with low turbidity and cloudless skies. Differences between AOD at 500 and 870 nm with PFR‐CIMEL, PFR‐PREDE, and PREDE‐CIMEL are shown in Fig.4, and Fig. 5. AOD shows a good agreement for CIMEL and PREDE at both channels, although the differences are slightly higher for channel 500 nm than 870 nm. The comparison between PREDE and CIMEL, for both 500 and 870 nm, showed a RMS of 0.005 and a bias of 0.003, below the nominal uncertainty of 0.01‐0.02 for both instruments. The comparison between PREDE and PFR for 870 nm showed a RMS of 0.006 and a bias of 0.005, whereas the results at 500 nm were 0.007 and 0.006, respectively. In the comparison between CIMEL and PFR, the RMS was 0.007 and 0.005 for channels 500 and 870 nm, with a bias of 0.005 and 0.003 respectively. The comparison between CIMEL and PREDE were within the WMO limits by better than 98% for both channels even of still occasional small exceedances are found. The agreement between PFR and CIMEL is 81% and 97% within the WMO limits for 500 nm and 870 nm respectively. Better filtering of the PFR data is necessary due to the contamination from shadows from the surroundings and clouds.

Figure 4: AOD differences versus time

Figure 5: AOD differences versus airmass; the green lines represent WMO recommended thresholds

5. Future plan for improving traceability

In this section will be briefly shown a plan for improving the traceability, and validating the ILP in situ calibration. This plan is based on the establishment of one “master” ESR/SKYNET instrument to be sent once a year to PMOD/WORCC for intercomparison campaigns against the PFR triad. Each campaign, that will last 10‐20 days depending on the cloud conditions, will be held in spring or fall. The “master” instrument, once back at its permanent site, will be used for calibrating two or three other ESR/SKYNET POMs, coming from permanent sites characterized by different environment, i.e. mountain, marine, urban, rural sites. The instruments involved in this campaign will be the “reference” instruments for the other sites of the network. Both the “master” and the “reference” instruments, once back to their sites, will perform simultaneously the monthly in situ calibration ILP. A good candidate as “master” ESR/SKYNET instrument is the one located at Monte Cimone, at about 2200 m a.s.l. (Fig 6) and managed by the Italian Air Force. This site could infact become a new high mountain laboratory where performing calibrations of the European POM’s.

Figure 6. The POM01 installed at the site of Monte Cimone

6. Conclusions

The intercomparisons, performed during the FRC‐IV Campaign and in Valencia, showed that AOD provided by POMs of ESR/SKYNET network are successfully comparable both against PFR‐GAW and CIMEL‐AERONET measurements, mainly at 870 nm. This result demonstrates that the in situ Improved Langley method, applied for calibrating the ESR/SKYNET network, is capable to monthly calibrate the instruments with accuracy. The technique can be applied directly at the site where the instrument is located, therefore reducing the expenses and time necessary by frequently shipping the instruments to calibration centers or high mountain sites, that are considered to be the more appropriate calibration locations. Moreover the methodology assures a more frequent check of the status of the equipment. A larger plan of intercomparison against WMO world reference PFRs at PMOD/WORCC and/or in some ESR/SKYNET sites has under development in order to: i) provide traceability of the AOD delivered by “masters” instruments of the network; ii) validate and test the in situ calibration procedures at environments with variable aerosol load. In this plan of traceability, the Italian ESR/SKYNET sites of the Italian Air Force will give a good contribution, both concerning the participation to intercomparison campaigns, and the implementation of the site of Monte Cimone, located at about 2200 m a.s.l., as a European high mountain laboratory where performing calibrations of ESR/SKYNET instruments and others.

7. Acknowledgments The participation of Víctor Estellés was financed by the Spanish Ministry of Economy and Competitiveness (MINECO) and the European Regional Development Fund through projects CGL2015‐64785‐R and CGL2015‐70432‐R, and by the Valencia Autonomous Government through projects PROMETEUII/2014/058 and GV/2014/046.

8. References

M. Campanelli, T. Nakajima, B. Olivieri, “Determination of the solar calibration constant for a sun sky radiometer. Proposal of an in situ procedure”. Applied Optics, Vol 43 n. 3, 20 January 2004.

M. Campanelli, V. Estellés, C. Tomasi, T. Nakajima, V. Malvestuto and J. A. Martínez‐Lozano. “Application of the SKYRAD improved Langley plot method for the in situ calibration of CIMEL sun‐skphotometers” Vol. 46, No. 14 May, 2007, Applied Optics.

M. Campanelli, et al., “Monitoring of Eyjafjallajoekull volcanic aerosol by the new European SkyRad users(ESR) sun‐sky radiometer network”, Vol 48, p 33‐45 , 2012, Atmospheric Environment

Holben et al., AERONET—A Federated Instrument Network and Data Archive for Aerosol Characterization, Remote Sensing of Environment, Volume 66, Issue 1, October 1998, Pages 1– 16doi:10.1016/S0034‐4257(98)00031‐5,1998

IPCC, 2013: Climate Change 2013: The Physical Science Basis. Cambridge University Press, Cambridge, United Kingdom and New York, NY, USA, 1535 pp,doi:10.1017/CBO9781107415324 Takamura, T., and T. Nakajima, Overview of SKYNET and its activities, Opt. Pura Apl. 37, 3303‐ 3308, 2004.

Wehrli, C., 2005. GAW‐PFR: a network of aerosol optical depth observations with precision filter radiometers. In: WMO/GAW Experts Workshop on a Global Surface Based Network for Long Term Observations of Column Aerosol Optical Properties Technical Report. GAW Report No. 162, WMO TD No. 1287.