Introduction
The World Meteorological Organization (WMO) through its Global Change
Observing System (GCOS) defined several essential climate variables (ECVs) as
physical, chemical or biological variables or a group of linked variables that
critically contributes to the characterization of Earth's climate. The ECVs
have been selected with the aim to obtain enough evidence that effectively
led us to predict the climate evolution and its possible associated risks.
(a) Location of the Izaña station (IZA) on a global map
of all BSRN stations (http://bsrn.awi.de, last access: 7 February 2019). (b) Izaña Atmospheric Observatory. Views of
Izaña radiation station: (c) northern and eastern views (azimuth
360∘, inclination 0∘; azimuth 90∘, inclination
0∘, respectively); (d) southern and western views (azimuth
180∘, inclination 0∘; azimuth 270∘, inclination
0∘, respectively).
Basic-BSRN radiation instruments installed between 2009 and 2017 at
IZA BSRN (SWD, DIR, DIF and LWD). The instruments currently in operation are
marked in bold. WRMC is
the World Radiation Monitoring Center.
Parameter
Manufacturer
Type
Serial
WRMC
Starting
Finish
Spectral
Instrument
number
date
date
range
uncertainty
SWD
Kipp & Zonen
CM-21
080034
61001
01/03/2009
10/11/2016
335–2600 nm
<±1 % for
EKO
MS-802F
F15509FR
61010
11/11/2016
–
285–3000 nm
daily totals
DIR
Kipp & Zonen
CH-1
080050
61003
01/03/2009
10/11/2016
200–4000 nm
<±1 % for
EKO
MS-56
F15048
61012
11/11/2016
–
daily totals
DIF
Kipp & Zonen
CM-21
080032
61002
01/03/2009
10/11/2016
335–2600 nm
<±1 % for
EKO
MS-802F
F15508FR
61011
11/11/2016
–
daily totals
080022
61004
01/03/2009
01/05/2009
050783
61008
01/05/2009
13/05/2014
080022
61004
14/05/2014
22/07/2014
<±3 % for
LWD
Kipp & Zonen
CGR-4
050783
61008
23/07/2014
30/03/2017
4.5–42 µm
daily totals
080022
61004
30/03/2017
07/06/2017
050783
61008
08/06/2017
–
Among others, the surface radiation budget and more specifically the surface
Earth radiation budget (ERB) longwave and surface ERB shortwave have been
identified as ECVs, due to their key role in the general circulation of the
atmosphere and ocean, and the thermal structure of the atmosphere, as well as being a
main factor in the Earth's climate system . The surface
ERB comprises the fluxes absorbed by the Earth's surface and the upward and
downward thermal radiative fluxes emitted by the surface and atmosphere,
respectively .
The first surface solar radiation measurements started in the 1920s in some
sites in Europe. The study of these historic datasets reveals an increase in
the surface solar radiation until the 1950s, known as “early
brightening” , but only observed in Europe due to the
scarcity of available data. The study of surface solar radiation long-term
records shows decadal changes with a decline of surface solar radiation from
the first available records, around 1950, until the middle of the 1980s
, a period known as “global dimming”, and an
increment in the surface solar radiation since the middle of 1980s, a period
known as “global brightening” .
All these studies remark on the variable quality of the data due to the
technical advances in the instruments since the 1970s; thus, the confidence of
the long-term trends observed should be taken into account when analyzing the
results. With the aim to obtain data with the best possible quality, in the
1990s, efforts were made to establish measurement networks around the Earth
with high-quality requirements to avoid introducing undesirable uncertainties
in the long-term series. In this context, the Baseline Surface Radiation
Network (BSRN) was proposed in 1980 by the WMO and created in 1992 to provide
accurate irradiances at selected sites around the Earth, with a high temporal
resolution. The BSRN is a project of the World Climate Research Programme
(WCRP). In 2004, it was designated as the baseline network for GCOS. The
available data cover a period from 1992 to the present thanks to the
contribution of 59 stations covering various climate zones
(http://bsrn.awi.de/nc/stations/maps/, last access: 7 February 2019). The BSRN data have been widely used, due to their quality and
reliability in the validation of satellite observations, as input to climate
models and to monitor the solar radiation reaching the Earth's surface
.
The BSRN imposes very strict measurement requirements in order to assure the
required quality of data . A BSRN site must be
representative of the surrounding area and avoid pollution sources, unnatural
reflectance, microclimate conditions and human activities that can affect
its representativity of the surroundings . Consequently,
BSRN sites cannot be located near major roadways, airports, vehicle
parking areas and buildings.
In 2009, the Izaña Atmospheric Observatory (IZA, BSRN station no. 61)
started its process to become a BSRN station through a specific agreement
between the State Meteorological Agency of Spain (AEMET) and the University
of Valladolid. IZA was proposed and accepted to be part of the BSRN at the
11th BSRN Workshop and Scientific Review meeting was held in Queenstown, New
Zealand, in August 2012 , and has since remained a member of the
network without interruption.
Between 2013 and 2014, the UV-A and UV-B radiation measurements performed at
IZA were used for satellite-based data validation. The validation resulted in
a good agreement with satellite-based data, and it is the starting point for
further developments of Flyby's elaboration processes .
The main goal of this work is to present the status of the Izaña BSRN
(IZA) between 2009 and 2017. Section describes the IZA site. The
main characteristics of the instruments and measurements that are part of IZA
BSRN as well as instrument calibrations are presented in
Sect. . Section illustrates data processing and
quality control procedures applied to the measurements, and the shipment
station-to-BSRN archive. Finally, a summary and conclusions are given in
Sect. .
Site description
IZA station (http://izana.aemet.es, last access: 7 February 2019) is managed by the Izaña Atmospheric Research Center (IARC) and is
part of AEMET. It is located on the island of Tenerife (Canary Islands,
Spain, at 28.3∘ N, 16.5∘ W; 2373 m a.s.l.)
(Fig. ).
Basic-BSRN radiation instruments currently in operation at IZA BSRN.
(a) SWD: EKO MS-802F pyranometer installed on a table (horizontal);
(b) DIF: EKO MS-802F pyranometer; (c) LWD: Kipp & Zonen
CGR4 pyrgeometer; (d) DIR: EKO MS-56 pyrheliometer; and
(e) Sun tracker: Owel INTRA 3.
Extended-BSRN radiation instruments installed between 2009 and 2017
at IZA BSRN (UV-B, UV-A, SWD and LWU). Same as
Table .
Parameter
Manufacturer
Type
Serial
WRMC
Starting
Finish
Spectral
Instrument
number
date
date
range
uncertainty
970839
61007
01/03/2009
22/02/2010
<±5 % for
UV-B
Yankee YES
UVB-1
071221
61009
22/02/2010
22/07/2015
280–315 nm
daily totals
970839
61007
23/07/2015
–
UV-A
Kipp & Zonen
UV-A-S-T
08005
61006
01/03/2009
–
315–400 nm
<±5 % for
daily totals
Kipp & Zonen
CRN1
030693
61005
01/03/2009
27/11/2016
PYR: 305–2800 nm
<10 % for
SWD
PYRG: 5–50 µm
PYRG and <5 %
LWD
EKO
MR-60
S15115.07
61013
01/01/2017
–
PYR: 285–3000 nm
for PYR
PYRG: 3–50 µm
daily totals
Extended-BSRN radiation instruments currently in operation at IZA
BSRN. (a) UV-B: Yankee YES pyranometer; (b) UV-A: Kipp &
Zonen UV-A-S-T pyranometer; and (c) SWU and LWU: EKO M-60 net
radiation sensor.
IZA is a high-mountain station above a quasi-permanent strong temperature
inversion layer that prevents the arrival of local pollution from lower
levels of the island. This meteorological feature favors measurements under
free troposphere conditions . As a result, the climate in the
area of the station is extremely dry for the majority of the year; this,
together with clean air from middle/upper troposphere, gives the area a
high scientific interest. Pristine conditions are alternated with periodical
intrusions of the dust-laden Saharan air layer
, mainly in summertime. IZA registers
the highest average annual insolation duration of Spain with about 3473 h of
sunshine per year and an average of 179.5 days per year of clear days in the climate period (1981–2010) (for more
information, see http://www.aemet.es, last access: 7 February 2019).
IZA enrolled in the WMO Global Atmosphere Watch (GAW) programme in 1989. In
addition, IZA has contributed to several international networks such as NDACC
(Network for the Detection of Atmospheric Composite Change;
http://www.ndsc.ncep.noaa.gov, last access: 7 February 2019) since 1999, and GAW-PFR (Precision Filter Radiometer Network;
http://www.pmodwrc.ch/worcc, last access: 7 February 2019)
since 2001. In 2003, the WMO/GAW Regional Brewer Calibration Centre for
Europe (RBCC-E; http://www.rbcc-e.org, last access: 7 February 2019) was established at IZA. IZA has been part of the Aerosol
Robotic Network (AERONET; http://aeronet.gsfc.nasa.gov, last access: 7 February 2019) since 2004, as one of the two absolute AERONET
calibration sites. IZA has also been a BSRN station since 2009 .
Moreover, in 2014, IZA was appointed by WMO as a CIMO (Commission for
Instruments and Methods of Observation) test bed for aerosols and water vapor
remote sensing instruments . A detailed description of the IZA
site and its observation programs can be found in .
Measurements and instruments
Basic-BSRN measurements
The basic-BSRN measurements of the BSRN program at IZA are global shortwave
radiation (SWD), direct radiation (DIR), diffuse radiation (DIF) and longwave
downward radiation (LWD) (Table ).
Other instruments currently providing data to IZA BSRN: (a)
radiosonde profiles, (b) the Brewer spectrophotometer and
(c) SONA camera (automatic cloud observation system) installed at
IZA and examples of images taken by the SONA camera.
Daily control test of the Owel INTRA solar tracker at IZA.
(a) Front view of the quadrants of the Sun detector. The numbers indicated are
pins of the sensor/connector, respectively; (b) Sun-sensor signals in the
four quadrants and the total signal from the quadrants; (c) board
temperature (∘C); and (d) current from base shunt of
motor0 driver (azimuth axis, mA) (red color) and current from base shunt of
motor1 driver (elevation axis, mA) (black color)
.
At present, SWD and DIF are measured with unshaded and shaded EKO MS-802F
pyranometers (Fig. a and b) (ISO-9060 classification: secondary
standard), respectively. This pyranometer is a high-precision instrument with
a spectral range from 285 to 3000 nm with a response time less than 5 s
(95 %, confidence level) and an expected uncertainty <±1 % for
daily totals. The BSRN accuracy target for DIF and SWD is 2 %
(5 W m-2) and 2 % (3 W m-2), respectively .
DIR is measured with an EKO MS-56 pyrheliometer (Fig. d) (ISO-9060
classification: first class). This instrument has a full operating view angle
of 5∘ and slope angle of 1∘. The spectral range covers from
200 to 4000 nm (50 % points) with a response time >1 s (95 %). The
expected uncertainty is <±1 % for daily totals. The BSRN accuracy
target for DIR is 0.5 % (1.5 W m-2) . LWD is
measured with a shaded Kipp & Zonen CGR4 pyranometer (Fig. c)
(ISO-9060 classification: secondary standard). The spectral range is
4.5–42 µm (50 % points) with a response time less than 6 s
(63 % response). The expected uncertainty is <3 % for daily totals
(95 %). The BSRN accuracy target for LWD is 2 % (3 W m-2)
.
Absolute cavity radiometer (PMO6) mounted on the IZA BSRN
Sun tracker during a calibration campaign.
Summary of calibrations of the different radiation instruments
performed at IZA between 2009 and 2017. PMOD
is the Physikalisch-Meteorologisches Observatorium Davos, and WRC is the World Radiation Center.
Parameter
Manufacturer
Serial number
Calibration date
Calibration site
Calibration factor
(µV W-1 m-2)
27/03/2008
Factory calibration
8.86
Kipp & Zonen
080034
Kipp & Zonen
SWD
CM-21
18/07/2014
IZA/AEMET
8.67
EKO
F15509FR
26/04/2016
Factory calibration
7.04
MS-802F
EKO
19/06/2008
Factory calibration
9.98
Kipp & Zonen
080050
Kipp & Zonen
DIR
CH-1
18/07/2014
IZA/AEMET
9.91
EKO
F15048
26/01/2016
Factory calibration
8.99
MS-56
EKO
27/03/2008
Factory calibration
8.63
Kipp & Zonen
080032
Kipp & Zonen
DIF
CM-21
18/07/2014
IZA/AEMET
8.68
EKO
F15509FR
26/04/2016
Factory calibration
7.05
MS-802F
EKO
28/02/2008
Factory calibration
10.37
080022
Kipp & Zonen
LWD
Kipp & Zonen
12/01/2007
Factory calibration
9.77
CGR-4
Kipp & Zonen
050783
30/06/2014
PMOD/WRC
9.39
27/03/2017
PMOD/WRC
9.41
UV-B
Yankee YES
071221
28/02/2008
Factory calibration
199*
UVB-1
Yankee YES
970839
23/08/2006
PMOD/WRC
0.1178*
12/07/2015
AEMET
0.1191*
UV-A
UV-A-S-T
27/10/2006
Factory calibration
32.469*
080005
Kipp & Zonen
08/08/2017
PMOD/WRC
30.98*
Kipp & Zonen
030693
08/01/2003
Factory calibration
10.09
Kipp & Zonen
SWU and LWU
EKO
S15115.07
12/05/2016
Factory calibration
SWU: 6.95; SWD: 6.8
MR-60
EKO
LWU: 3.08; LWD: 3.22
* (W m-2) V-1.
These instruments are installed on a Sun tracker, with the exception of the EKO MS-802F
pyranometers for SWD and DIF measurements, which are installed on a horizontal
table (Fig. a). The Sun tracker is an Owel INTRA 3
(Fig. e). This is an intelligent tracker which combines the
advantages of automatic-tracking operation (automatic alignment with the
system of astronomical coordinates) and actively controlled tracking (a
four-quadrant Sun sensor). It is constructed for use under extreme weather
conditions; its operational temperature range is between -20 and
50 ∘C. It can sustain about 50 kg of carefully balanced load. The
tracker motors have a special grease for use in low temperatures. It moves
back to the start (morning) position at the corresponding midnight. The drive
unit has a zenith rotation >90∘. The unit has an angular
resolution ≤0.1∘, an angular repeatability of ≤∼0.05∘ and an angular velocity ≥1.5∘ s-1 on the
outgoing shafts. The maximum speed is 2.42∘ s-1
.
Flowchart of data from measurements to the BSRN
database.
The lower and upper limits for the physically possible limits (PPLs) and
extremely rare limits (ERLs) used in flagging the radiation measurements.
µo is the cosine of the solar zenith angle and Sa is the solar
constant adjusted for the Earth–Sun distance.
Lower bound (W m-2)
Upper bound (W m-2)
Parameter
PPL
ERL
PPL
ERL
SWD
-4
-2
Sa×1.5×µo1.2+100
Sa×1.2×µo1.2+50
DIF
-4
-2
Sa×0.95×µo1.2+50
Sa×0.75×µo1.2+30
DIR
-4
-2
Sa
Sa×0.95×µo0.2+10
SWU
-4
-4
Sa×1.2×µo1.2+50
Sa×µo1.2+50
LWD
40
60
700
500
LWU
40
60
900
700
In addition, the measurements of pressure (P), relative humidity (RH) and
temperature (T) are included in this measurement group. The pressure is
measured a with Setra 470 pressure transducer, and RH and T are measured
with Campbell Scientific CS215-L sensors.
Extended-BSRN measurements
The extended-BSRN measurements included in the IZA BSRN program are shortwave
upward radiation (SWU), ultraviolet measurements (UV-A and UV-B) and
longwave upward radiation (LWU) (Table ).
Example of data-removing process. The rejected data were caused by
shadows during cleaning operations. The dashed line represents the
pyrheliometer cleaning time (a) before removing rejected data and
(b) after removing rejected data.
A Yankee YES pyranometer (Fig. a) measures global radiation in the
UV-B spectral range from 280 to 315 nm with a response time around 100 ms.
The UV-A (315–400 nm) is measured with a Kipp & Zonen UV-A-S-T pyranometer
(Fig. b) with a response time less than 1.5 s (95 %). The
expected uncertainty is <5 % for daily totals (95 %). SWU and LWU are
measured with a MS-60 EKO radiometer (Fig. c) (ISO-9060
classification: secondary standard). This system is formed by two
pyranometers and two pyrgeometers. The spectral range of the pyranometers is
280–3000 nm with a response time ∼17 s, while the spectral range of
the pyrgeometers is 3–50 µm.
The radiation measurements are acquired with a Campbell CR5000. This
data logger is a rugged, high-performance data-acquisition system with
a built-in keyboard, graphics display and PCMCIA card slot. It combines a 16-bit
resolution with a maximum of 5000 measurements per second. In particular, the
measurements are taken with a time step of 5 s. The minimum, average,
maximum and standard deviation are stored every minute.
Same as Table except for the “comparison” intervals used
for flagging the radiation quantities. σ is the Stefan–Boltzmann constant
(5.67×10-8 W m-2 K4), SZA is solar zenith angle, T
is air temperature (K), and SumSWD is DIR × cos(SZA) +
DIF.
Comparison
Conditions
Test
SZA <75∘
SWD/SumSWD ∼±8 %
SumSWD >50 W m-2
SWD/SumSWD
75∘< SZA <93∘
SWD/SumSWD ∼±15 %
SumSWD >50 W m-2
SumSWD <50 W m-2
Test not possible
SZA <75∘
DIF/SWD <1.05
SumSWD >50 W m-2
DIF/SWD
75∘< SZA <93∘
DIF/SWD <1.10
SumSWD >50 W m-2
SWD <50 W m-2
Test not possible
LWD to air temperature
0.4σT4< LWD <σT4+25 W m-2
Ancillary measurements
Ancillary measurements, such as radiosonde data (Fig. a) and total
ozone column (TOC), are performed at IZA BSRN station.
Vertical profiles of pressure, temperature, relative humidity and wind
direction and speed are measured using Vaisala RS92 radiosondes
that are launched twice a day, at 00:00 and
12:00 UTC at the Güimar station (WMO GRUAN station no. 60018;
105 m a.s.l.), managed by the Meteorological Centre of Santa Cruz de
Tenerife (AEMET). This station is located near the coastline at a distance in
a straight line from IZA of ∼15 km. The TOC measurements are performed
with the Brewer spectrophotometer (Fig. b) (precision better than
1 %) . An automatic cloud observation system (SONA
camera) (Fig. c) developed by Sieltec Canarias
S.L. takes all-sky images every 5 min, day and night. This camera consists
of a resolution of 640×480 pixels and an 8-bit color response CCD
sensor with a Bayer filter, with a spectral range from 400 to 700 nm. A
rotating shadow band is used for protecting the sensor from direct sunlight.
Instrument checks and maintenance
All the instruments of the BSRN are checked on a daily basis by
meteorological observers of the Izaña observatory. Routine checks consist
of cleaning the domes, cable connections inspection and instrument leveling,
as well as checking the proper functioning of the solar tracker and shading
system of the instruments for DIF and LWD measurements. Recently, a tool to
test the Owel INTRA solar tracker real-time check-up has been implemented.
This test consists in controlling the four quadrants of the solar tracker (see
Sect. ), checking its leveling, the board temperature and the
intensity of the base shunt of motor in azimuth and elevation axis
(Fig. ).
Percentage of rejected data (red color) and accepted data (blue
color) according to the PPLs and ERLs, and comparison of various irradiance components
(SWD, SumSWD, DIF, LWD and air temperature) at IZA between 2009 and 2017.
Instrument calibrations
All the radiation instruments (Tables and ) have been
periodically calibrated following the recommendations of the BSRN
(Table ) and are regularly compared with reference instruments with
recent calibration from the World Radiation Center (WRC) at Davos.
An absolute cavity pyrheliometer PMO6 designed at PMOD
(Physikalisch-Meteorologisches Observatorium Davos) (Fig. ) that is
regularly calibrated at the WRC is used as a reference instrument and is
directly traceable to the World Radiometric Reference. Periodical
calibrations with PMO6 allow us to assure the reliability of the measurements
and correct time degradation on the calibration constants. A large
calibration campaign of BSRN pyranometers and pyrheliometers was performed
during 2014 using the aforementioned PMO6. The ISO 9059:1990(E) and ISO
9846:1993(E) recommendations were met. The calibration of a field
pyranometer/pyrheliometer by means of a reference pyrheliometer is
accomplished by exposing the two instruments to the same solar radiation and
comparing their corresponding measurements. This allows us to compare target
instruments to high-accuracy radiation sensors. A second calibration campaign
was held in July–August 2018.
BSRN IZA management
During 2009, a BSRN database was developed in order to manage the large volume
of BSRN data. This tool not only allows the management of a large volume of
information that is automatically generated, but it is also used for checking of
real-time measurements and becoming a comprehensive quality control system with
corresponding alarms.
The BSRN data management flowchart is shown in Fig. . It includes
daily and monthly semi-automatic processes to collect and check the
measurements, and generate the station-to-archive file sent every month to
the BSRN database. This daily process can be executed automatically, or on
demand, producing several warning alerts if human checking is needed.
Data are stored in a CR5000 data logger (see Sect. ). This
data logger generates a raw data file that is stored in a database
for further analysis, if necessary, which is also available on an internal
web for real-time access. The raw data file is checked in order to
assess the format integrity and detect gaps.
The input parameters of the LibRadtran model, their sources and
corresponding references.
Input
Source
Input parameters
References
measured at IZA
Radiative transfer
Disort2
–
equation solver
Atmosphere model
Long-term ozonesonde performed
x
at the Botanic Observatory
(BTO; Tenerife)
Solar flux
Kurucz
–
Ozone cross section
Bass and Paur
–
Absorption parameterization
SBDART
–
Surface albedo
0.11
x
Ozone column
Brewer spectrophotometer
x
Water vapor column
AERONET products
x
Aerosol Angstrom
AERONET products
x
Aerosol asymmetry albedo
AERONET products
x
Aerosol single
AERONET products
x
scattering albedo
Aerosol profile
Shettle
–
Number of streams
16
–
–
Scatterplots and histograms of the instantaneous (W m-2) and
daily (MJ m-2) radiation measurements and simulations for the period
2009–2017: (a, d, g, j) SWD, (b, e, h, k) DIR and
(c, f, i, l) DIF. The fitting parameters are shown in the
legend.
Measurement radiation corrections
Several corrections are applied to raw data to obtain the final radiation
data. These corrections are as follows:
Zero offset: This is defined as the signal caused by changes in the instrument temperature. The zero offset is
measured for each instrument as part of the observation sequence when possible. For instruments that are not capable
of obtaining a zero offset with each observation, it is measured at night and subtracted from daytime values .
The average values of zero offset compared to the radiation values performed during the day are rather small,
representing 0.3 % and 0.02 % of the SWD and DIR signals for 1000 W m-2, respectively .
Cleaning operations: As remarked in Sect. , daily cleaning of domes is performed. Some artificial shadows are
caused when the observers perform these operations. Data corresponding to cleaning activities are identified and removed from the database (Fig. ).
Exceptional situations: Shadows or gaps in raw data are also observed due to exceptional situations, such as severe weather,
repairing of instruments and maintenance operations, etc. Data stored during these non-operational periods are also removed from the database.
Quality control (QC)
Once the corrections are made, a Dep data file is obtained, which will be used
to perform the quality control (QC) tests. The IZA QC procedure has two main
parts: the recommended BSRN controls and the comparison with simulations
with radiative transfer models (RTMs).
The first part of the QC consists of applying the QC methods that the WRMC
recommends to the BSRN data .
These quality control procedures are based on checking whether the
measurements are within certain limits: physically possible limits (PPLs),
extremely rare limits (ERLs) and the comparison of various irradiance
components.
The PPL procedure is introduced for detecting extremely large errors in
radiation data, while the ERL procedure is used to identify measurements
exceeding the extremely rare limit. Radiation data exceeding these limits
normally occur under very rare conditions and over very short time periods.
These tests are based on empirical relations of different quantities
(Table ).
The final BSRN QC procedure is the comparison of various radiation
components, i.e., the ratio between the DIR, directly measured with a
pyrheliometer, and the derived value from the difference between the SWD and
DIF (SumSWD), and the ratio between DIF and SWD. These tests capture smaller
errors that have not been detected by the PPL and ERL procedures
(Table ).
Statistics for the bias between instantaneous and daily SWD, DIR and
DIF simulations and measurements at IZA BSRN for the
period 2009–2017. MB is mean bias; SD is standard deviation; RMSE is root mean square error.
MB (%)
SD (%)
RMSE (%)
Instantaneous
Daily
Instantaneous
Daily
Instantaneous
Daily
SWD
-1.68
-1.24
2.26
1.03
2.28
1.58
DIR
-1.57
-1.82
1.92
1.17
2.00
2.07
DIF
0.08
0.84
7.90
8.69
9.89
9.11
At IZA, the measurements' quality assessment is performed taking into account
the tests described above, by using the BSRN Toolbox software
developed for the BSRN community and WRMC. This
software also includes a data format check for the station to archive files
and for PANGAEA download files (see below). Data quality checks as outlined
in the BSRN global network recommended QC tests V2.0 can
also be performed with this software. We have found that <1 % of all
radiation measurements at IZA are outside the PPL and ERL limits (see
Fig. a and b) between 2009 and 2017 for solar zenith angles
(SZAs) <90∘.
The ratio between the different components also confirms the high quality of
the SWD, DIR and DIF measurements. For SWD/SumSWD and SZA <75∘, >98 % (Fig. c) of the data are between 0.92 and
1.08, while for 75∘ < SZA <93∘, 96 % of the
measurements range from 0.85 to 1.15 (Fig. d). For DIF/SWD, the
results present a high quality with 99 % within the established limits, for
both SZA <75∘ and 75∘ < SZA < 93∘
(Fig. e and f). The IZA radiation measurements largely meet the
BSRN quality controls.
BSRN Izaña station home page
(http://www.bsrn.aemet.es, last access: 7 February 2019).
BSRN Izaña: BSRN data description, QC test (BSRN global network
recommended QC for SWD, DIR and DIF) and BSRN–model comparison for SWD, DIR,
DIF and UV-B radiation at IZA using the LibRadtran RTM
(http://www.bsrn.aemet.es, last access: 7 February 2019).
The second part of the QC is the comparison of instantaneous and daily
radiation measurements with simulations performed with RTMs during clear
periods. An adaption of Long and Ackerman's method for
IZA, performed by , is used for detecting instantaneous
clear-sky periods. This method is based on 1 min SWD and DIF measurements to
which four individual tests are applied to normalized SWD, maximum DIF,
change in SWD with time and normalized DIF ratio variability.
Following the BSRN recommendations, the instantaneous clear-sky periods
detected are simulated and compared with instantaneous and daily radiation
measurements. The RTM model used is LibRadtran
(http://www.libradtran.org, last access: 7 February 2019;
), which has been extensively tested at IZA
. The measured input parameters used in the
LibRadtran model simulations are shown in Table . The model input
parameters – precipitable water vapor (PWV), aerosol optical depth (AOD),
total ozone column and surface albedo – are measured at IZA
. The straightforward comparison between the
instantaneous and daily SWD, DIR and DIF simulations and measurements shows
an excellent agreement (Fig. ). The variance of daily
(instantaneous) measurements overall agrees within 99 % (98 %) of the
variance of daily (instantaneous) simulations.
The simulations slightly underestimate the instantaneous/daily measurements
of SWD (-1.68 %/-1.24 %) and DIR (-1.57 %/-1.82 %), while the
DIF simulations overestimate the instantaneous/daily measurements
(0.08 %/0.84 %). The RMSE is <2.5 % for SWD and DIR for both
instantaneous and daily comparisons, while for DIF it increases to 9.89 %
and 9.11 % for instantaneous and daily comparisons, respectively
(Table ). These results are in agreement with those obtained by
.
Web tool
With the aim to have, at a glance, an overview of the state of the BSRN
station, a web site has been developed for the IZA BSRN station
(Fig. ; http://www.bsrn.aemet.es, last access: 7 February 2019). Plots for several variables such as SWD, DIR, DIF, UV-B
and UVI index are automatically available at the home web page, as well as
corresponding simulations performed with actual input data at the IZA
station. These plots are provided in near-real time (every 10 min).
IZA time series of daily SWD, DIR, DIF, UV-A and UV-B
measurements, and instantaneous LWD data series (11:00 UTC) for the period
2009–2017.
On the web page, there are links to the comparison between measurements and
simulations, QC control results, long-term series and derived products, among
other additional information. Additional information on the installed
instrumentation and the BSRN-related publications is also available
(Fig. ). In the following paragraphs, we present a short
description of the BSRN Izaña, long-term series and derived products.
BSRN Izaña: In this menu, it is possible to select any date and plot the results of applying the QC recommended by the BSRN
(see Sect. ) for SWD, DIR and DIF. It is also possible to plot the comparison of measured and simulated SWD, DIR, DIF and UV-B
radiation at IZA using LibRadtran RTM and input parameters measured at IZA. This section of the web is automatically updated every night,
once the measured input parameters for the model are available, and the QC tests are applied, according to the flowchart shown in Fig. .
Long-term series: Daily values of SWD, DIR, DIF, UV-A and UV-B measurements and instantaneous (11:00 UTC) values of LWD are correspondingly
updated on the web (Fig. ). SWD measurements started in 1977 with a bimetallic pyranometer (PYR) being replaced in
1992 by different instruments (Kipp & Zonen: CM-5, CM-11 and CM-21) .
Derived products: From DIR measurements and following the methodology developed by , the apparent transmission is automatically
calculated for the purposes of monitoring clear-sky solar transmission (Fig. ). This apparent transmission is defined as the ratio of the output
from a normal-incidence pyrheliometer for a specific pair of SZA corresponding to integer air masses on the morning of a given day:
τ=(Idir/ITOA×sinh)1/ma,
where Idir is DIR, ITOA is the top of the atmosphere (TOA) irradiance, ma is absolute air mass and h is solar elevation angle.
Data time series of direct solar radiation atmospheric transmission determined at IZA for the period 2009–2017.
Station-to-BSRN archive file
The last step in the IZA data management is to create the station-to-archive
file, which is submitted to the BSRN database on a monthly basis. This
procedure is performed using the radiation measurements (Sect.
and ), radiosonde profiles and total ozone data
(Sect. ).
Example of (a) visualization of daily data of SWD, DIR, DIF
and LWD radiation at IZA on August 2017 (software: PanPlot;
) and (b) intercomparison between SWD on the
x axis and SumSWD (defined as the sum of DIF and DIR on a horizontal plane) on
the y axis calculated using the BSRN Toolbox
.
As shown in Fig. , the process includes the application of QC tests
again to the radiation data using the specific software tools developed by
the BSRN. A visual inspection of the monthly data series is made to avoid
outliers or detect erroneous data of the different variables before
submitting the file to the BSRN database (Fig. ). Finally, if the
checks are correct, the station-to-archive file is sent by ftp
(http://ftp.bsrn.awi.de, last access: 7 February 2019).
Scientific exploitation of IZA BSRN data
A detailed description of the BSRN database has recently been published by
. In this paper, some end-user applications of the IZA BSRN
data are described.
The IZA BSRN data have been used in diverse research works encompassing
several research fields. The most recent publications that have used the IZA
BSRN data as part of their work are listed in Table , grouped by
the research field. Here, we only remark on the peer-reviewed works but it
should be noted that there are many contributions and proceedings that are
related also to a greater or lesser extent to the IZA BSRN data.
Publications that used the IZA BSRN data in the last
5 years.
Research field
References
Solar spectra determination
Testing of new photocatalytic materials
Study of ultrafine particle formation processes
Reconstruction of historical data series
Comparison with simulations from RTMs
Neural networks
and multi-regression techniques
Estimating the underlying effective albedo
Satellite measurements'/forecasts' validation
Climate and long-term studies
Solar energy application
In future research, the IZA BSRN data will be essential to accurately analyze
the attenuation of different types of clouds in UV, visible and infrared
radiation, and to study the optical and radiative properties of mineral dust,
as well as for solar energy applications, such as solar radiation nowcasting.
Conclusions
The Izaña station has been part of the BSRN since 2009. IZA BSRN
contributes with basic-BSRN radiation measurements: global shortwave
radiation (SWD), direct radiation (DIR), diffuse radiation (DIF) and longwave
downward radiation (LWD); extended-BSRN measurements: ultraviolet radiation
(UV-A and UV-B), shortwave upward radiation (SWU) and longwave upward
radiation (LWU); and other measurements: vertical profiles of temperature,
humidity and wind obtained from radiosonde profiles (WMO station no. 60018) and total
ozone column thickness from the Brewer spectrophotometer.
Following the recommendations of the BSRN, the quality control tests have
been routinely applied. The analysis of the QC results shows very good data
quality that meets the BSRN requirements. The percentage of measurements that
are outside the PPLs and ERLs is <1 % for SZA <90∘ in the period 2009–2017. The
ratios between components also provide good results, with >98 % of
measurements within the limits for SZA <75∘. The poorest result
is the SWD/SumSWD for 75∘ < SZA <93∘ with >96 %
of measurements within the defined limits.
In addition, we have compared the instantaneous and daily SWD, DIR and DIF
measurements with simulations obtained with the LibRadtran RTM. The observed
agreement between measurements and simulations is very good: the variance of
daily and instantaneous measurements overall agrees within 99 % and 98 %,
respectively. The simulations underestimate the instantaneous/daily
measurements of SWD (-1.68 %/-1.24 %) and DIR (-1.57 %/-1.82 %), while DIF simulations overestimate the instantaneous/daily
measurements (0.08 %/0.84 %). The RMSE is lower than 2.5 % for SWD
and DIF for both instantaneous and daily comparisons. These results
demonstrated a high consistency between the measurements and simulations
reinforcing the reported data quality. The results show also the usefulness
of the RTM as a tool for quality control radiation measurements over time.