doi: 10.7289/V5N29TWP

GSICS Quarterly: Special Issue on Ultraviolet

duced during the on-ground characterization, since under ambient conditions
it is very difficult to perfectly thermally
stabilize and remove (through normalization) the etalon introduced by water
layers at the detector level.

Volume 8, No. 2, 2014

References
Slijkhuis, S., Wahl, S., Aberle, B.,
and Loyola, D., 2004, Re-analysis of
GOME/ERS-2 diffuser properties,
Envisat & SRS Symposium, Salzburg,
Austria.

GOME-2 Factsheet, EUM/OPS/
DOC/10/1299, Version 4A
GOME-2/Metop-A Level 1B Product
Validation Report No. 5, Version 1F
Rate and comment on this article

Use of Solar Reference Spectra for Satellite
Instruments
by Matthew DeLand, SSAI
The extraterrestrial solar irradiance (i.e.,
measured outside Earth’s atmosphere)
has numerous benefits as a calibration source for remote sensing satellite
instruments. For example:
• It is available for an extended period
of time during every orbit.
• Its intensity is high across a wide
spectral range.
• It is, for much of the spectrum,
extremely stable compared to any
onboard calibration source. And, in
regions where it is not stable, the
variations are often highly correlated.
• Many spectral features are available
for use in wavelength calibration.
Each of these benefits also brings
some challenge in terms of instrument
design and data analysis:
• The constant presence of solar illumination requires instrument designers
to limit exposure of optical surfaces to
minimize degradation, particularly at
UV wavelengths.
• Solar irradiance can be stronger than
terrestrial radiance at a given wavelength by factors ranging from 102 to
104, which must be considered when
designing the optical system and
instrument electronics.
• Solar irradiance does have natural
variability at UV wavelengths (particularly below 300 nm), with time
scales ranging from minutes to years.
• The solar spectrum incorporates
millions of emission and absorption
features, so that the exact irradiance
Return to Page 1

spectrum produced by any instrument
will depend on its characteristics
and design.
In order to use the Sun as a calibration source, it is necessary to first have
a reference spectrum to understand
what the instrument is expected to see.
This article will briefly describe some
published reference solar spectra, and issues that users should be aware of when
selecting a dataset for their own use.
Overview
It is important for users to realize that
no single instrument measures solar irradiance at high resolution simultaneously
over all spectral regions of interest (Xray, UV, visible, IR). Thus, any reference
solar spectrum is typically a composite
of measurements from multiple instruments, taken at different times, and
often with varying spectral resolution.
In addition, the final reference spectrum
may incorporate radiometric adjustments based on comparisons with lower
resolution measurements. Differences
in spectral resolution and sampling are
particularly important in the UV, where
many deep Fraunhofer lines occur. The
next section presents basic information about some reference solar spectra
published during the last decade. Table 1
summarizes specific parameters for each
spectrum.
Reference Solar Spectra
ATLAS. The solar reference spec-

tra created by Thuillier et al. (2004)
take advantage of the ATLAS missions
conducted in the early 1990s, when
three solar irradiance instruments (along
with other remote sensing instruments)
were flown together on the NASA Space
Shuttle. The UARS satellite (carrying
two additional solar instruments) was
also operating during this period, so
that up to five datasets were available
for some spectral regions. Thuillier et
al. selected reference dates during the
ATLAS-1 and ATLAS-3 missions (29
March 1992 and 11 November 1994,
respectively) to construct spectra corresponding to moderately high and
moderately low levels of solar activity.
Each spectrum covers the spectral range
0.5-2,400 nm, using rocket data for the
EUV region below 120 nm. They created average irradiance values for UV
spectral regions where multiple data sets
were present. They also used the high
resolution model spectrum of Kurucz
(1995) to insert spectral features into
lower-resolution measurements in the
visible and IR regions.
KNMI. Dobber et al. (2008) created a
high resolution solar reference spectrum
covering 250–550 nm to support onorbit calibration of the OMI instrument
on the EOS Aura satellite. Very high
resolution data from AFGL balloon measurements (Hall and Anderson, 1991)
and KPNO ground-based measurements
(Kurucz et al., 1984) were convolved
to a slightly lower resolution, and
then adjusted radiometrically through
6

doi: 10.7289/V5N29TWP

GSICS Quarterly: Special Issue on Ultraviolet

Volume 8, No. 2, 2014

Table 1. Summary of Solar Reference Spectra
Name
WvL
Sampling
ATLAS
0.5–2400 nm
0.5–120 nm: 1 nm
120–400 nm: 0.05
nm
400–2400 nm:
0.2–0.6 nm
KNMI
250–550 nm

Sources

Resolution

Time

Calibration

Accuracy

0.5-120 nm: Rocket
120-200 nm: UARS
SOLSTICE, UARS SUSIM
200-400 nm: UARS
SOLSTICE, UARS
SUSIM, ATLAS SUSIM,
SSBUV, SOLSPEC
400-870 nm: SOLSPEC
870-2400 nm: SOSP

0.5-120 nm:
1.0 nm
120-400 nm: 0.25
nm (smoothed)
400–2,400 nm:
0.5 nm (degraded
model)

0.5-120 nm: 1991 rocket
120–2400 nm: 29 Mar
1992 for ATLAS-1, 11
Nov 1994 for ATLAS-3

Satellite data +
normalization of
integrated TSI

0.5–120 nm: 30–40%
120–2,400 nm: 2–4%

200–310 nm: AFGL
balloon
300–550 nm: KPNO

0.025 nm
(convolved)

Multiple dates

Adjusted based on
comparison with
low resolution data
(UARS SUSIM)

< 5% for 300–550 nm

0.1-6 nm: TIMED XPS
6–106 nm: MEGS EVE
(rocket)
106-116 nm: TIMED SEE
116–310 nm: SORCE
SOLSTICE
310–2400 nm: SORCE
SIM

0.1–310 nm: 0.1 nm
310–2400 nm: 1–30
nm

Quiet: 10–16 Apr 2008
Active: 25–29 Mar 2008,
29 Mar – 4 Apr 2008

Satellite data +
normalization of
integrated TSI to
SORCE TIM

0.1-116 nm: 10–15%
116–2,400 nm: 2–4%

200–300 nm: AFGL
300–1,000 nm: KPNO

0.04 nm
(convolved)

Multiple dates

Adjusted based on
comparison with
low resolution data
(ATLAS-1)

< 5% for 300–1,000
nm

0.01 nm
WHI
0.1–2400 nm
0.01 nm

SAO 2010
200–1000 nm
0.01 nm

comparisons with UARS SUSIM data
(Floyd et al., 2003) and balloon data
(Gurlit et al., 2005). It should be noted
that any terrestrial measurements of absolute solar irradiance must also account
for absorption effects due to the Earth’s
atmosphere.
WHI. During the most recent solar
minimum between Cycles 23 and 24,
a focused program called the Whole
Heliosphere Interval (WHI) collected
approximately simultaneous solar
measurements at minimum activity
conditions from X-ray to IR wavelengths. Woods et al. (2009) used rocket
measurements to supplement operational
satellite measurements over a wide
spectral range (0.1–2,400 nm) and produce three reference spectra representing active and quiet conditions during
March-April 2008. The primary data
sources are EVE rocket data in the X-ray
and EUV regions, SORCE SOLSTICE
data in the FUV and MUV, and SORCE
SIM data from the NUV to the IR. The
Return to Page 1

only radiometric adjustment came from
comparison of the integrated reference
spectrum to SORCE TIM total solar irradiance data.
SAO 2010. Chance and Kurucz
(2010) have also produced a solar reference spectrum using the AFGL balloon
data and the KPNO ground-based data,
in this case covering the spectral range
200–1000 nm. They used a revised
analysis of the KPNO data, and normalized those data to the Thuillier et al.
(2004) ATLAS-1 spectrum longward of
300 nm.
Effects of Resolution and Sampling
Differences in spectral resolution,
either from the original measurements or
from later re-convolution, and in dataset
sampling should be considered when
selecting a solar reference spectrum. To
illustrate this point, Figure 1 shows irradiance data between 300–320 nm from
each of the reference spectra described
here. While the KNMI and SAO 2010

spectra are both based on KPNO data
and sampled at 0.01 nm, the slightly
broader bandpass used for SAO 2010 is
evident in Figure 1(b). The WHI spectrum shown in Figure 1(d) is reported at
0.1 nm sampling over its entire range,
but the change in resolution between
SOLSTICE and SIM data at 310 nm
is clear.
Conclusion
Numerous solar irradiance reference
spectra, created by combining multiple
data sets, are currently available. These
spectra can differ in spectral resolution
and sampling by a factor of 10. The
quoted accuracy of each spectrum is
typically 2–4% at UV and visible wavelengths, but differences between spectra
can show larger variations locally.
Choosing a specific reference spectrum
for regular use should be guided by user
requirements.

7

doi: 10.7289/V5N29TWP

GSICS Quarterly: Special Issue on Ultraviolet

Volume 8, No. 2, 2014

Floyd, L., Rottman, G., DeLand, M., and
Pap, J., 2003, Eleven years of solar UV
irradiance measurements from UARS.
Proc. ISCS 2003 Symposium “Solar
Variability as an Input to the Earth’s
Environment, SP-535, ESA, Noordwijk,
p. 195-203.
Gurlit, W., Bösch, H., Bovensmann, H.,
Burrows, J.P., Butz, A., Camy-Peyret,
C., Dorf, M., Gerilowski, K., Lindner,
A., Noël, S., Platt, U., Weidner, F., and
Pfeilsticker, K., 2005, The UV-A and
visible solar irradiance spectrum: intercomparison of absolutely calibrated,
spectrally medium resolution solar
irradiance spectra from balloon- and
satellite-borne measurements. Atmos.
Chem. Phys., 5, 1879–1890.
Hall, L. A., and Anderson, G.P., 1991,
High resolution solar spectrum between
2000 and 3100 Å. J. Geophys. Res., 96,
12,927–12,931.
Kurucz, R.L., 1995, in Proc. 17th Annual Conference Transmission Models
(ed. G. P. Anderson), p. 333.
Kurucz, R. L., Furenhild, I., Brault, J.,
and Testermann, L., 1984, Solar flux atlas from 296 to 1300 nm, National Solar
Observatory Atlas No. 1, Sunspot, NM.
Thuillier, G., Floyd, L., Woods, T.N.,
Cebula, R., Hilsenrath, E. Hersé, M., and
Labs, D., 2004, Solar irradiance reference spectra, in Solar Variability and its
Effects on Climate, AGU Monograph
141 (ed. J. Pap), p. 171–193.

Figure 1. Sample data at 300–320 nm from different solar reference spectra. (a) = KNMI, (b) = SAO 2010, (c) =
ATLAS-1, (d) = WHI. Res = resolution, Samp = sampling.

References
Chance, K., and Kurucz, R. L., 2010, An
improved high resolution solar reference spectrum for earth’s atmosphere
measurements in the ultraviolet, visible,
and near infrared. J. Quant. Spect. Rad.
Trans., 111, 1289–1295.

Return to Page 1

Dobber, M., Voors, R., Dirksen, R.,
Kleipool, Q., and Levelt, P., 2008, The
high resolution solar reference spectrum
between 250 and 550 nm and its application to measurements with the Ozone
Monitoring Instrument. Solar Phys.,
249, 281–291.

Woods, T. N., Chamberlin, P.C., Harder,
J.W., Hock, R.A., Snow, M., Eparvier,
F.G., Fontenla, J., McClintock, W.E.,
and Richard, E.C., 2009, Solar Irradiance Reference Spectra (SIRS) for
the 2008 Whole Heliosphere Interval
(WHI), Geophys. Res. Lett., 36, L01101,
doi:10.1029/2008GL036373.

Rate and comment on this article

8