Climate Station Info
Measurement instruments
The solar radiation measurement instruments are from the Dutch company Kipp and Zonen B.V. and comprise two pyranometers for measuring shortwave solar irradiance, one pyrheliometer for measuring shortwave direct normal irradiance and a pyrgeometer for measuring longwave downwelling radiation.
The solar radiation measurement instruments are mounted on a tracker from Kipp and Zonen, type SOLYS 2. The tracker accurately points the pyrheliometer and the shading ball assembly mounted on the tracker at the sun. A sun sensor mounted on the tracker fine tunes the tracking of the sun.
A shading ball assembly screen of the beam irradiance from one of the two pyranometers and from the pyrgeometer. The view angle from the pyrheliometer to the sky is 5° and the view angle from the pyranometer to the shadow ball is 5°. In this way, the pyranometer measures exactly the amount of solar irradiance that is not screened of by the shadow ball while the pyrheliometer measures exactly the amount of solar irradiance that is screened of by the shadow ball.
The shading ball screen of the beam irradiance from the pyrgeometer in order to protect the infrared window of the pyrgeometer from heating up, and creating an offset in the readings.
The other meteorological parameters are measured by a weather transmitter, type WXT520 is from the Finnish company Vaisala Oyj.
Table 1 shows the irradiance measurement instruments used at the climate station while Table 2 shows the other meteorological parameters from WXT520 used at the climate station. For wind measurements, scalar averaging is used for both wind speed and direction.
Table 1: Solar radiation measurement equipment.
Measurement

Sensor type

Resolution

Accuracy

Global shortwave solar irradiance

CMP 11

0.1 W/m^{2}

1.4 %

Short wave horizontal solar irradiance

CMP 11

0.1 W/m^{2}

1.4 %

Short wave direct normal solar irradiance

CHP 1

0.1 W/m^{2}

1 %

Downwelling horizontal longwave irradiance

CGR 4

0.1 W/m^{2}

< 1 %

Table 2: Other meteorological parameters.

Range

Resolution

Accuracy

Wind direction minimum [°]

0  360°
(The direction from where the wind blows. 0°=North, 90°=East, 180°=South, 270°=Vest)

1°

3°

Wind direction average [°]

Wind direction maximum [°]

Wind speed minimum [m/s]

0 – 60 m/s

0.1 m/s

The greater of 0.3 m/s or 3% of the measurement range 0…35 m/s
5% for the measurement range of 36…60 m/s

Wind speed average [m/s]

Wind speed maximum [m/s]

Air temperature [°C]

52 – 60 °C

0.1 °C

0.2 K at ambient temperature of 50 °C – 0 °C
0.3 K at ambient temperature of 20 °C.
0.4 K at ambient temperature of 40 °C
0.7 K at ambient temperature of 60 °C

Relative humidity [%]

0 – 100 %RH

0.1 %RH

3 %RH at 0…90 %RH
5 %RH at 90…100 %RH

Air pressure [hPa]

600 – 1100 hPa

0.1 hPa

1 hPa at ambient temperature range 52…+60 °C

Rain fall [mm]

Collecting area: 60 cm^{2}

0.01 mm

Cumulative accumulation after latest auto reset.
< 5 %, weather dependent. Due to the nature of the phenomenon, deviations caused by spatial variations may exist in precipitation readings, especially in short time scale. The accuracy specification does not include possible wind induced error.

Rain duration [s]


10 s

Counting each 10second increment whenever droplets detected.

Rain intensity [mm/h]

0 – 200 mm/h


Running one minute average in 10second steps.

Hail [hits/cm^{2}]


0.1 hits/cm^{2}

Cumulative amount of hits against collecting surface.

Hail duration [s]


10 s

Counting each 10second increment whenever hailstone detected.

Hail intensity [hits/m^{2}/h]


10 hits/cm^{2}/h

Oneminute running average in 10second steps.

Quality control and gap filling of the measured climate data
In order to ensure the data quality of the measured weather data, the measurement instruments are cleaned and calibrated on a regular basis. The measurements are filtered for physical impossible values and possible gaps are filled before the data is distributed.
The measured weather data are distributed as:
 filtered data where data gaps are filled (BSRN filtered data – gaps filled)
 filtered data (BSRN filtered data – no gaps filled)
 raw measured data where data gaps are filled (Non filtered data – gaps filled)
 raw measured data (Non filtered data – no gaps filled)
Filtering the data to eliminate physical impossible data
The filtering to detect physical impossible data is done by applying a filter in accordance with the Baseline Surface Radiation Network (BSRN) standard. Physical impossible data are deleted from the filtered files and replaced in the gap filling procedure.
BSRN is a network under the World Climate Research Program (WCRP) (Ohmura et al. 1998) aiming to provide the climate community with accurate and highly resolved irradiances for climate research purposes. This global network measures surface radiative fluxes at the highest possible accuracy with well calibrated stateoftheart instrumentation at selected sites in major climate zones, (Roesch et al. 2011).
The filters for physically possible intervals on horizontal are listed in table 3. S_{0} is the solar constant adjusted for EarthSun distance, µ is the cosine of the solar zenith angle, GLOB is the global irradiance, SWDIFF is the shortwave horizontal diffuse irradiance, SWDIR is the shortwave direct normal irradiance, LWDOWN is the downwelling longwave irradiance. Irradiance values that fall outside the physically possible irradiance intervals are deleted and the gap is filled by the gap filling procedure described in the next section.
Table 3: Upper and lower limits for the physically possible irradiance intervals on horizontal, (Roesch et al. 2011).
Parameter

Lower limit

Upper limit

GLOB

4 W/m^{2}

1.5·S_{0}·µ^{1.2}+100 W/m^{2}

SWDIFF

4 W/m^{2}

0.95·S_{0}·µ^{1.2}+100 W/m^{2}

SWDIR

4 W/m^{2}

S_{0}·µ^{1.2}

LWDOWN

40 W/m^{2}

700 W/m^{2}

The solar constant adjusted for EarthSun distance S_{0} is calculated by the approximation (1)
S_{0} = 1367 W/m^{2} · (1+0.033·cos(360·n/365)) (1)
The cosine of the solar zenith angle, µ is calculated by equation (2)
µ = cos(SZA) (2)
Where
n = the day number. January 1 is day number 1.
SZA= solar zenith angle, the angle between the vertical and the line to the sun [°]
The calculation of the zenith angle, SZA is explained in (Jakob: Her vil vi gerne have et link til undervisningsnotat om solstråling. Notatet er ikke offentligt tilgængeligt, men vi har en pdf fil som kan ligge hvor det passer dig).
Gap filling procedure
Data gaps caused by missing or false data are filled by the method described by (Schwandt et al. 2013, Skartveit et al. 1998).
The solar radiation gaps are filled either by using equation relation of the three solar radiation components, modeled values, linear interpolation of clearness indices or by replacing data from neighboring days.
The relationship between the three solar radiation components i.e. global horizontal irradiance (GHI), shortwave horizontal diffuse irradiance (DHI) and direct normal irradiance (DNI) is described by equation (3). The modeled values are the clearness index (k) and the diffuse fraction (d) and the values are calculated by means of the Skartveit Model (Skartveit et al. 1998). The relationship between the modeled values and the solar radiation components are described by the equations (4) and (5).
GHI = DHI + DNI·µ (3)
k=GHI/(S_{0}·µ) (4)
d=DHI/GHI (5)
The gap filling procedure distinguishes between missing solar radiation data and missing other meteorological data. For solar radiation data, the procedure distinguishes between:
 The availability of three components of solar radiation being i.e. GHI, DHI and DNI
 The length of the gap i.e. gaps less than 3 hours and gaps greater than 3 hours and gaps greater than 24 hours.
This leads to three cases:
Case 1: One component is missing.
 The missing component is found by equation (3)
Case 2: Both DHI and DNI are missing.
 GHI is used as input to the Skartveit Model (Skartveit et al. 1998) to calculate clearness index, k and diffuse fraction, d. Finally, the missing DHI and DNI values are calculated by (5) and (3) respectively.
Case 3: All three solar irradiance components are missing.
 If the data gap is less than 3 hours, the clearness indices are calculated for GHI and DNI for all the time steps where GHI and DNI are available on both sides of the gap. The clearness indices, k for the gap are then calculated by linear interpolation and used as input to the Skartveit Model (Skartveit et al. 1998) to calculate the diffuse fraction, d. Finally, GHI and DHI values are calculated by (4) and (5) respectively and DNI values are calculated by (3).
 If the data gap is larger than 3 hours, the gap is filled with data from neighboring days if data is available. The gap filling procedure can fill data gaps up to 10 days if data before and after the gap are available. The first 5 days will be replaced by data from the day before the gap and the last 5 days with data from the day after end of the gap. The limit of 10 days is set due to the fact that in atmospheric science it is assumed that weather stays constant for a period of 5 days. Further, since the sun position does not deviate significantly for a period of 5 days, this procedure is applied to solar radiation data so that the data matches with sunrise and sunset times and also with sun elevation and azimuth angles (Schwandt et al. 2013).
The gap filling procedure for other meteorological data follows the method described in case 3. Instead of linear interpolation of the clearness indices, the values of the parameters are directly linear interpolated.
Data resolution
The data resolution can be selected to 1 minute, 30 minutes, 1 hour or 24 hours.
Data resolution of 1 minute contains the mean value of the measurement in the previous minute. For rain and hail, the measurements contain the sum measured in the previous minute. Maximum and minimum values are the maximum and minimum values measured in the previous minute.
For data resolution of 30 minutes, 1 hour and 24 hours, the average value or the sum of the previous time span is used. In this way, the time 30 minutes contains the average value or the sum from the time span 0 minutes to 30 minutes, and so on.
References
I. Reda, A. Andreas, ‘Solar Position Algorithm for Solar Radiation Applications’, National Renewable Energy Laboratory, NREL. NREL/TP56034302, 2008.
M. Schwandt, K. Chhatbar, R. Meyer, K. Fross, I. Mitra, R. Vashistha, G. Giridhar, S. Gomathinayagam, A. Kumar, ‘Development and test of gap filling procedures for solar radiation data of the Indian SRRA measurement network’, ISES Solar World Congress, 2013.
A. Skartveit, J. Olseth, and M. E. Tuft, ‘An Hourly Diffuse Fraction Model With Correction For Variability And Surface Albedo’, Solar Energy, vol. 63, no. 3, pp. 171–183, 1998.
A. Ohmura, E. Dutton, B. Forgan, C. Frölich, H. Gilgen, H. Hegner, A. Heimo, G. KönigLanglo, B. Mcarthur, G. Müller, R. Philipona, R. Pinker, C. Whitlock, M. Wild, ‘Baseline Surface Radiation Network (BSRN/WRMC), a new precision radiometry for climate research’, B. Am. Meteorol. Soc., 79, 2115–2136, 1998.
A. Roesch, M. Wild, A. Ohmura, E.G. Dutton, C. N. Long, T. Zhang, ‘Assessment of BSRN radiation records for computation of monthly means’ Atmos. Meas. Tech., 4, 339354, 2011.