1. INTRODUCTION

The measurement of solar spectral irradiance (SSI: spectrally resolved power of the solar electromagnetic radiation in $\mathrm{W}{\mathrm{m}}^{-2}{\mathrm{nm}}^{-1}$) from soft x-ray/extreme ultraviolet (XUV/EUV) to infrared (IR) and total solar irradiance (TSI: integrated power of the solar electromagnetic radiation in $\mathrm{W}{\mathrm{m}}^{-2}$) provides the basic data for solar, solar-terrestrial, atmospheric, and climate physics. SSI data of higher absolute accuracy will support activities in these fields of research.

Solar physics implies the knowledge of photon fluxes excited in the solar atmospheric altitude regions, i.e., the photosphere, the lower and upper chromosphere, the transition region and the corona, with plasma temperatures increasing with altitude from about 6000 K to $>{10}^6\mathrm{K}$. SSI photons are therefore strongly dependent on the altitude region in which they are generated. The coverage of the photons extends from the IR to the XUV spectral regions to be measured at a cadence from minutes (flare observation) and days (solar rotation) to 11 years (solar cycle) at the highest possible spectral resolution and with absolute accuracy. Solar modeling is key for different aspects. It can also be used to check and improve SSI accuracy in spectral regions of weak solar energy output. When SSI data are measured and collected by instruments with different spectral resolutions, solar modeling can provide a good tool for the interpretation of the SSI data, especially when it comes to integrating SSI and comparing it with TSI.

Solar-terrestrial and atmospheric physics investigates the SSI spectral regions being absorbed in the terrestrial atmosphere. The thermosphere-ionosphere region and, to a certain extent, the upper mesospheric altitude regions are controlled by XUV and EUV photons. The EUV is absorbed in the upper atmosphere, leading to the partial ionization of its atoms and molecules and the subsequent heating of the thermosphere. A change in EUV leads to changes in electron density, temperature, and neutral density [1,2]. The ionosphere plays an important role in radio communication too. EUV also controls the dissociation of molecular oxygen (${\mathrm{O}}_2$) and carbon dioxide (${\mathrm{CO}}_2$), which means that it plays a dominant role in the thermosphere-ionosphere region and partially in the upper mesosphere down to about 65 km. Application of occultation, e.g., in the EUV, provides a good way of measuring thermospheric densities and temperatures, including exospheric temperatures [3]. Following the significant decrease of these parameters during the past solar minimum (2008–2009) with decreased XUV/EUV fluxes, [4,5] future measurements are of great interest in order to find out whether this unusual decline continues. The SSI of the vacuum ultraviolet (VUV) and IR ranges plays a major role in the upper mesosphere down to the tropospheric interaction with ozone and other atmospheric trace gases as well as with aerosols. Absolute SSI data at medium spectral resolution on the basis of a daily average make an important contribution to this area of research. SSI and TSI measurements are also input into planetary atmosphere models to validate these models by comparing their output with measurements to allow determination of the most accurate results.

Climate research calls for the highest possible absolute accuracy of SSI and TSI data on solar cycle periods at medium spectral resolution and data sets averaged over solar rotations of about 28 days each. For climate modeling, the total spectral range of SSI data is required in order to account for vertical and horizontal energy exchange from the bottom to the thermospheric-ionospheric system of the Earth’s atmosphere. Improving the accuracy of SSI data for climate modeling remains a major issue in the VUV spectral region, where the absorption of SSI by atmospheric trace gases initiates long chains of physical-chemical processes that still require further investigation. In this context, the primary anthropogenic trace gas driving climate change is carbon dioxide. Methane (${\mathrm{CH}}_4$) and chlorofluorocarbon (CFC) species that cause depletion of stratospheric ozone (${\mathrm{O}}_3$) also play a role. A special phenomenon is observed above the tropopause. There, the atmosphere has cooled during recent solar cycles to an extent that exceeds the cooling caused by the decreasing EUV SSI. As atmospheric density decreases with altitude, the IR bands enable trace gases such as ${\mathrm{CO}}_2$ to absorb and emit radiation, becoming optically thinner such that radiative cooling begins to dominate over absorption and warming [4].

Applying SSI data to the scientific areas mentioned, the requirements significantly differ from each other. In addition, the knowledge of the SSI variation must be taken into account for a complete understanding of the effects on the atmosphere and ultimately on climate change [6]. SSI variability is a function of wavelength [7]. While the solar XUV/EUV output varies by $>100\%$ during a solar cycle, SSI variability decreases almost uniformly to $\approx 0.15\%$ in the visible (VIS) spectral region.

There are two more factors that add to the difficulty of deriving accurate SSI data: namely, the progressive degradation of SPs in space and the lack of spectral irradiance sources that closely simulate the SSI over the full spectral range with respect to energetic and spectral composition as well as to angular divergence.

Precise measurements of SSI can only be achieved through repeated calibration of the SPs in order to exclude the impact of degradation on the spectrometric recordings, which is a challenging task [8].

2. BASIC REQUIREMENTS FOR CALIBRATING SSI INSTRUMENTATION

The correction of SSI observations using onboard tracking of space instrument degradation has been a difficult task to date [9,10]. In addition, there is no standard radiation source available that allows satisfactory simulation of the natural SSI. Even worse, the manifold effects of the space environment on SSI SPs cause instrument degradation and means that laboratory calibration data can only be considered provisional. Furthermore, the possibilities for implementing calibration in satellites are quite limited. Therefore, degradation of SSI instrumentation in space poses a challenge for the community. Moreover, degradation leads to high uncertainties. To track the degradation, previous missions have used either ionization chambers (ICs) as a primary irradiance detector [11,12], internal lamps [13], and backup channels [14] as an onboard calibration source, or stars as noncalibrated relatively stable radiation sources to determine changes in the sensitivity of the SPs in the VIS spectral range [15]. This publication presents an instrument design and measuring method that aims to substantially improve the accuracy of SSI data acquisition. First, requirements are derived for comparing SSI space instruments and methods in order to optimize degradation tracking or to exclude the impact of degradation on recorded spectrometric data.

A. SSI Instruments and Causes of Degradation

From 1975 to 1978, a Committee on Space Research (COSPAR) Task Force supported by a group of 31 participants investigated the impact of degradation on the SSI data derived from SPs operating in space. The only way to separate SSI variability caused by solar activity from changes in instrumental efficiency caused by degradation is “to perform solar flux monitoring from long mission satellites with recalibration measurements” [16]. This conclusion was based on the observation that all types of SSI SPs considered (Ebert–Fastie, Rowland circle, Czerny–Turner, Seya–Namioka-SPs, and others) suffer from degradation that is challenging to monitor in space because too many different processes are involved. Usually SSI itself with photon energies $>100\mathrm{eV}$ leads to the greatest degradation. Cosmic radiation, contaminants such as hydrocarbons, temperature changes, very low gas pressures, and other factors can change the properties of optical surfaces, filters, and detectors. This can result in an increase or decrease in the spectrometric efficiency, which can change with wavelength [17]. Though different methods to correct for degradation have been applied, the results still require substantial improvements.

B. Calibration in Space: Requirements for SSI Measurements

Two methods are typically used to quantify degradation or to exclude the impact of degradation on instruments: namely, calibration of SPs with (i) standard radiation sources to simulate the SSI and (ii) primary irradiance detectors.

1. Calibration with Standard Radiation Sources

To date, SPs as mentioned in Section 2.A have been used. For absolute calibration, the geometry of the entrance slit or aperture is the fundamental parameter for calibrating the SP. It fixes the critical “sun-imaged effective area” on the first optical component of the SP, controlling its efficiency; its areal size defines the number of photons available for the correct determination of the SSI. In this context, the following requirements $R_n$ are derived:

In satellites, standard radiation sources are not mounted in front of the SSI instruments, meaning that the radiation does not pass the entrance slit. Instead, the sources are mounted inside the SP system. The photons reach the SP along a path that differs from that of the SSI. Such SPs therefore do not meet the requirements:

In the laboratory, these types of SPs are precalibrated with high expenditure. In space, the large equipment used is not applicable. While the simulated radiation of the laboratory sources passes the entrance slit of the SPs, the standard radiation generated by lamps in the satellite instruments take other paths and also interact with other optical surface areas and components of different efficiencies. Since the optical path of the radiation from the standard source differs from the one of the SSI, degradation strongly differs between the two radiation paths.

This means that SPs with an entrance slit do not meet the top priority requirements for accurate calibration in space.

2. Calibration Using Primary Irradiance Detectors

This calibration method refers to the scheme published in 1976 [18]. First, to determine the absolute photon flux, EUV irradiance is allowed to pass through a monochromator and enter an IC. Exchanging the IC with a SP allows one to determine its efficiency. In the solar autocalibrating EUV spectrometer system (SolACES), the monochromator is exchanged with a series of filters [19]. In order to cover the full SSI spectral regions, SOLACER also includes TSI primary irradiance detectors of the Davos Absolute RAdiometer (DARA) type. This method of SSI calibration in space is checked against these requirements:

The new SSI system SOLACER meets the requirements (see also Figs. 1 and 2 and the description of the measuring system).

 

Fig. 1. Scheme of the SOLACER subsystems (see also Section 3.A): The SP (green rectangle on the left) consists of a planar grating, a parabolic mirror, and a detector. Spectral recording is carried out while the grating turns. When SSI enters the SP through the open aperture (white rectangle in the filter wheel on the left below SSI), the signal $S_{\mathrm{SSI}}(\lambda )$ is generated. When a filter (red rectangle in the filter wheel on the left below SSI) is turned around and placed on the top SSI SP, the corresponding signal is $S_{\mathrm{SSIF}}(\lambda )$. The detectors are operated with filters. For cross-calibrating the DARA detectors with outer TSI devices or for cross-checking the BOS detectors, these detectors are operated without filters.

Download Full Size | PDF

 

Fig. 2. Scheme of the SOLACER subsystems (top view).

Download Full Size | PDF

3. SOLACER: A NEW IN-ORBIT CALIBRATING SSI MEASURING SYSTEM

TSI detectors are considered long-term stable primary irradiance detectors that perform with an accuracy of close to $2.2*{10}^{-4}$ [20,21]. If the TSI could be subdivided into two spectral ranges using two optical filters with known transmission characteristics, a first approach of SSI measurement capability could be to allocate two absolute spectral irradiance values in $\mathrm{W}{\mathrm{m}}^{-2}$ transmitted by the filters to both spectral subranges of TSI each. SOLACER’s goal is to subdivide TSI into an optimized number of subranges by filters covering the full spectral coverage of the solar electromagnetic emissions for deriving the absolute irradiance values for each spectral subrange of the filters. SSI recording by SPs also allows the consecutive determination of the actual filter transmission. In the XUV/EUV spectral regions, ICs serve as primary irradiance detectors.

For the XUV/EUV spectral regions, SolACES aboard the International Space Station validated the detector-based calibration method during nine years of operation [11,12]. For SOLACER, the spectral range of SolACES is extended to the IR by integrating TSI detectors.

The method for onboard degradation tracking is based on repeated in-orbit calibration. It is executed in five steps (see also Fig. 1 and Section 4):

A. SOLACER Optical Arrangement

Figure 2 schematically shows the combination of the subsystems:

All these elements are placed behind a rotatable filter wheel with two rings, each with 25 apertures of equal size. Two attenuator positions are assigned to the apertures of the PMTs in order to check their linearity and to adapt the counting rates to optimum operational conditions. The attenuators consist of three metallic nondegrading meshes of 10%, 1%, and 0.1% transparency, respectively (not shown in Fig. 2).

The BOS detectors add to the DARAs and to the PMTs. For the spectral range $<175\mathrm{nm}$, PMTs with different photon entrance windows (calcium/magnesium fluoride, thin quartz plates) and other photocathode materials are available [22].

More details on the main components and calibration are presented in the following sections.

B. SOLACER SPs

The planar grating geometry of SPs with one detector is chosen for its high transparency of radiation, the compact design [23–25], and the smooth photon efficiency function. Array detectors would require the complex determination of the efficiency function for each individual detector.

Nearly parallel solar radiation enters the SPs via a circular aperture with an 8 mm diameter, reflected and diffracted by the planar grating, and focused by a one-dimensional parabolic mirror onto the exit slit with the detector behind. The SSI spectra are scanned by either turning the grating or turning the fixed mirror and detector around the center of the grating, parallel to the grooves/lines.

Changing the aperture size, e.g., up to $1{\mathrm{cm}}^2$ or more does not change the spectral resolution determined by the focal length of the parabolic mirror or the angular divergence of the incoming beam. The size of the aperture is one of the parameters that can be adapted to optimize the accuracy of the SOLACER data.

High transparency: The aperture for SOLACER is $A_{\text{SOLACER}}=50{\mathrm{mm}}^2$. The slit area of the conventional SP airglow-solar spectrometer instrument (ASSI) [22] with a curved grating was $A_{\mathrm{ASSI}}=3{\mathrm{mm}}^2$, which means the difference was $\sim 17$. Count rates of SolACES ranged from ${10}^5$ to almost ${10}^8$ cps (counts per second). The expected count rates for PMTs are listed in Table 2, and the expected signal-to-noise ratio (SNR) for the SPs equipped with PbS detectors is shown in Table 3.

Compactness: The eight SPs are packed in four sets of two (Fig. 2). The size of the SolACES double-SPs is $85\mathrm{mm}*190\mathrm{mm}*320\mathrm{mm}$. The complete SOLACER instrumentation is estimated at about $480\mathrm{mm}*420\mathrm{mm}*350\mathrm{mm}$.

The typical spectral resolution is shown in Table 1. There are spectrometric parameters such as the grating constant and the focal length of the parabolic mirror to adapt the spectral resolution to other requirements.

Table 1. SP Properties: All Planar Gratings Have the Same Geometrical Size, While the Number of Lines/mm and the Coating Materials Differ

View Table | View all tables in this article

Table 1 also shows other characteristics of the planar grating SPs.

Spectral recordings take about 6 min each. During up to 16 orbits per day, there is sufficient time to record more spectra per day, which improves the SSI data statistics.

With reference to Fig. 2, SPs SP1 to SP4, the IC, one radiometer DARA, one BOS, and one PMT are located behind the inner ring of the filter wheel. SPs SP5 to SP8 and the other components, DARA, BOS, and PMT are operated via the outer circle. The doubled SPs SP3 and SP5 as well as SP4 and SP6 connect the corresponding spectral ranges with each other and provide the cross-calibration capabilities of the different calibration detectors. They take measurements in the lower SSI energy ranges to increase data accuracy.

C. SOLACER ICs

Figure 3 shows a diagram with a filter on the entrance of the double IC with front and rear electrodes and a photodiode at the end of the optical path of the filtered ${\mathrm{SSI}}_F$. Ion-electron pairs originating from the absorption of XUV/EUV photons in the ionization areas generate currents. The amplitude and curvature of the three currents depend on the gas pressure and photoabsorption cross sections $\sigma (\lambda )$. For most calibration cycles, the absorbing gas increases from 0 to 2 mbar, at which pressure all photons are absorbed. Then, adding up the currents from the front and rear electrodes, $I_f$ and $I_r$, and taking into account that one photon produces one ion-electron pair in the EUV region with $\lambda >25\mathrm{nm}$ (Fig. 4), the number of photons ${\mathrm{m}}^{-2}{\mathrm{s}}^{-1}\mathrm{\Phi }(\mathrm{\Delta }{\lambda }_F)$ is

 

Fig. 3. Scheme of the IC and a graph of currents generated in the front and rear ionization area and in the photodiode.

Download Full Size | PDF

 

Fig. 4. Primary and secondary production of ion-electron pairs for different absorbing gases in the IC.

Download Full Size | PDF

The filters are pressed onto the entrance aperture, sealing the IC. After a measurement period of approximately 5 min, the filters are taken off and the gas is released into space. Four cylindrical gas reservoirs with a volume of $510{\mathrm{cm}}^3$ each are also part of the instrumental structure. To ensure the astronauts’ safety, the maximum permissible gas pressure is 5.6 bar. After nine years of extensive calibration activities with the ICs, there was still sufficient gas available for another extension of the mission. In satellites, the maximum permissible gas pressure is up to 200 bar, which—if required for very long-term missions—can be used with increased volumes of the gas reservoirs.

The current from the photodiode is used for cross-checking purposes. Contrary to the other currents, it is not affected by secondary absorption.

Based on the experience and results of the SOLAR SolACES mission, we expect stable performance for a mission duration extending a solar cycle. Applied improvements derived from measurements at the Berlin Electron Synchrotron (BESSY) cover the spectral range from 2 to 150 nm [26] in the SOLACER SP system. Calibration capability and stability are based on the nonchanging photoabsorption cross sections. The production of ion-electron pairs in space is the same as in the laboratory.

XUV/EUV SSI uncertainties of 7.5% were observed during the SolACES mission, as explained in detail in [27]. Eight sources of uncertainties are taken into account. Given the very strong decrease of the spectrometric count rates after two years of International Space Station (ISS) SolACES operation (the nominal mission period was 18 months, with a subsequent extension to nine years). Five percent of the uncertainties are allocated to the corresponding low count rate statistics. This figure is expected to improve by $<1\%$ with SOLACER to $<5\%$ of uncertainties. State-of-the-art in this spectral range is 10% accuracy of SSI data. In view of the variability exceeding 100% and partially even 1 order of magnitude in the XUV/EUV, good progress has already been achieved with SolACES. It will significantly be improved with SOLACER.

In the spectral range of 25–150 nm, one EUV photon absorbed by gases such as helium, neon, argon, xenon, or nitric oxide produces one ion-electron pair each (Fig. 4). With this single production of ion-electron pairs, the spectral subregions extend from $\sim 25$ to 150 nm. Below 25 nm, the secondary ion-electron contribution adds to $I_f$ and $I_r$.

D. TSI-DARA Detectors

SOLACER is equipped with two TSI instruments for the conversion of the photon fluxes passing the filters into spectral irradiance ($\mathrm{W}{\mathrm{m}}^{-2}$). They are of the DARA-TSI type, which is an update of Compact and Light-weight Absolute Radiometer (CLARA), flown on the NORSAT-1 satellite as described in [28].

TSI space radiometers are well-known for their high long-term stability and sensitivity across the full solar spectrum. The integration of TSI detectors, ICs, BOSs, and PMTs in SOLACER provides a full-scale autocalibration instrument to derive highly accurate XUV/EUV/VUV/VIS/IR SSI data.

The DARA-TSI uses the electrical substitution principle to measure solar irradiance at the position of the satellite. The operation modes and data retrieval algorithms that were developed for the TSI instrument CLARA on NORSAT-1 are also implemented here. The onboard software actively controls the electric heating of the measuring cavity to maintain its thermal equilibrium with the so-called compensating cavity, which is continuously heated with a constant electrical current (Fig. 5). The third cavity serves as a backup channel. A mechanical shutter is operated every 15 s to alternately shade and irradiate the measuring cavity. The difference in electrical heater power between the shaded and irradiated states is then equal to the incident solar irradiance. The cavity absorptivity and area of the entrance aperture are determined before launch.

 

Fig. 5. Left panel, three cavity detectors from DARA mounted to a common heat sink. The cavities are painted black on the inside and gold on the outside. The solar irradiance is absorbed inside the cavities (irradiance originates from the top right corner); adapted from [28]. Right panel, DARA working principle. (a) The reference cavity is always shielded from sunlight and heated electrically ($P_{\text{Reference}}$) via a constant power of $\sim 45\mathrm{mW}$. The temperatures in the three cavities (TNC, TRC, and TBC) and the heat sink (THS) are constantly measured. The control circuit continuously adjusts the electrical heater power in the nominal and backup cavities to maintain the heat flux QNC and QBC identical to that of the reference cavity (QRC). (b) To measure irradiance, the shutter in front of the nominal cavity is opened, and the absorbed sunlight results in the cavity additionally being optically heated, so that less electrical power is required to keep the thermal flux QNC identical to QRC. The optical power in the nominal cavity is then equal to the difference between the electrical heater power in the closed ($P_{\text{closed}}$) and the open ($P_{\mathrm{open}}$) shutter states. The solar irradiance is then defined as $IS=(P_{\text{closed}}-P_{\mathrm{open}})/A$, with A being the aperture area.

Download Full Size | PDF

The absorptivity of the DARA can be considered independent of wavelength. An identical transfer filter is used for both the DARA absolute radiometer and the IC measurement. With DARA, we will determine the reference irradiance value for each bandpass of the transfer filters. This value will then be used to absolutely calibrate the spectra measured with the respective transfer filters in the UV, VIS, and NIR.

The cavity absorptivity is expected to slowly degrade when exposed to hard XUV-EUV-UV radiation in space. The cavity of the backup channel is only occasionally used to monitor the degradation process of the measuring cavity. The backup cavity remains shaded most of the time and only measures TSI during the so-called backup mode, simultaneously with the measuring cavity. After a certain number of backup operations, the compensating cavity is used to determine the degradation of the backup channel. During this mode of operation, the measuring cavity takes on the role of compensating cavity.

Because the requirement on the accuracy of the DARA detectors located in SOLACER can be released, the size of the current aperture measuring will be increased. Then cross-calibration with a separate TSI instrument aboard or outside of the same payload will be applied.

E. Secondary Irradiance Detectors: BOS and PMT Devices

Below 300 nm, the SSI is too weak to generate sufficiently accurate DARA signals. This is why SOLACER will also be equipped with two BOSs derived from the space-qualified sensor BOS [29] setup on the PICARD satellite and PMT devices. Their higher sensitivities will extend the DARA sensitivity range into the VUV and EUV spectral ranges (see also Table 4).

BOS is a broadband radiation measurement “instrument.” The main detector is a thermistor attached to a black-coated surface, which is permanently exposed to space without any optical and aperture accessories. The temperature measurements are used within a transfer function to determine variations in incoming solar irradiance as well as the terrestrial radiation.

The measurement principle of the BOS is a combination of a thermistor-based BOS and a calorimeter. The basic geometry is a cylinder, where the heat flux measurement is carried out through temperature measurements $T_1$ and $T_2$ at thermal nodes (Fig. 6).

 

Fig. 6. Schematic layout of the original BOS setup on the PICARD satellite.

Download Full Size | PDF

The sensing unit is composed of two aluminum masses: $m_1$ and $m_2$. The surface of $m_1$ is painted black; $m_2$’s surface is painted white. The mass $m_1$ is then connected to $m_2$ via a thin shunt at the bottom of $m_2$. In addition, $m_1$ is thermally insulated from $m_2$ with a multilayer insulation.

The smaller mass $m_1$ absorbs and re-emits electromagnetic radiation of a large bandwidth from space. The larger mass $m_2$ absorbs less energy (due to the white coating) and re-emits most of the energy caught by $m_1$ and $m_2$ back into space.

The working principle of measuring the incident radiation on BOS is as follows:

During exposure (i.e., absorption and re-emission) of the electromagnetic radiation, a temperature difference between $m_1$ and $m_2$ is observed. The incident radiation flux can be calculated using several material constants and the calculation of the thermal energy of this network.

The device is expected to provide SSI signals with a resolution of $0.1\mathrm{W}{\mathrm{m}}^{-2}$ and a dynamic range ${\mathrm{DR}}_{\mathrm{BOS}}={10}^6$. Since BOS is not operated on an absolute scale, it is cross-calibrated against the DARAs.

Two PMTs are used as secondary irradiance detectors. As an example, the Hamamatsu PMT R13096 is selected. It has a dynamic range of $1/9.5·{10}^6$, with a spectral response of 10%–90% from 185 to 900 nm. Extending the spectral range towards shorter wavelengths, a calcium or magnesium fluoride window will be chosen.

In Table 2, the expected counting rates (cps) of the Hamamatsu PMT are shown for the bridged spectral region only. For wavelengths between 300 and 175 nm, the photon energy (Energy), the SSI data, the corresponding number of photons (Photons), the optical transmission (OT) passing the optical gratings and parabolic mirrors are taken into account. The OT data are published [19]. PMT efficiencies are shown in the Hamamatsu data sheet.

Table 2. PMT Estimated Counts per Second for Hamamatsu PMT R13096

View Table | View all tables in this article

Table 3. SNR Calculation for a PbS-IR Detector of a Generic $\mathsf{D}*=\mathsf{1.5}\times {\mathsf{10}}^{\mathsf{9}}\mathsf{cm}{\mathsf{Hz}}^{\mathsf{1}/\mathsf{2}}/\mathsf{W}$ and an Integration Time of 1 s and an Area of $\mathsf{50}{\mathsf{mm}}^{\mathsf{2}}$, Yielding a noise equivalent power (NEP) of $\mathsf{4.7}\times {\mathsf{10}}^{-\mathsf{10}}\mathsf{W}$

View Table | View all tables in this article

The SSI variability in the IR spectral range $>800\mathrm{nm}$ is in the order of $\approx 0.15\%$. A correct SSI measurement requires at least 0.05% accuracy, which is not possible with PbS detectors with reasonable expenditure. Therefore, PbS detectors of the SPs are cooled with Peltier elements. Tracing the transmission of IR filters allows a more accurate measurement using TSI DARA and BOS detectors.

4. CALIBRATION

Using any source emitting electromagnetic radiation similar to the Sun, SOLACER is capable of self-calibration. TSI DARA detectors and ICs convert the incoming photon fluxes (photons ${\mathrm{m}}^{-2}{\mathrm{s}}^{-1}$) via the filters (Fig. 1) into spectral irradiance ($\mathrm{W}{\mathrm{m}}^{-2}$). For a description of the process, please refer to Section 3.

A. Calibration in the Laboratory and from the Ground

In the laboratory complex, instrumentation of standard radiation sources is used that is not available in space. On ground, the instruments must be tested and qualified in order to bring the detectors to SSI levels, to prepare the detailed operational procedures, and to develop software that can be applied during the mission to evaluate the recorded data. The BESSY electron synchrotron, the Physikalisch-Meteorologisches Observatorium Davos (PMOD)/World Radiation Center (WRC), Physikalisch-Technische Bundesanstalt (PTB) Braunschweig Black Body radiator, and the Fraunhofer Institute for Physical Measurement Techniques (IPM) laboratories with SSI simulation lamps are best suited for such activities.

In order to determine preliminary smooth efficiency curves ${\eta }_L=f(\lambda )$ of the SPs, SOLACER is irradiated directly with monochromatic BESSY radiation ${\mathrm{SI}}_L(\lambda )$ producing spectral signals $S_{\mathrm{SIL}}(\lambda )$. This is done for each of the eight spectrometric wavelength ranges in accordance with the scheme shown in Fig. 1 and the description in Section 3,

Stray light as well as the higher orders of radiation are determined on the basis of recordings of monochromatic radiation. This makes full profiles for both parameters with their respective wavelengths available, which can be used to correct their contributions to the recorded spectra. Filters fulfill the same role throughout the spectral regions of interest. This is a big advantage of this measuring method over other SSI SPs, which do not use filters.

Applying continuum radiation, data evaluation software is developed using the filters and defining their bandwidths to achieve proper SNR for the TSI DARA and BOS detectors. The spectral range from about 150 to 250 nm is key for achieving first-approach efficiency data.

The calibration parameters derived in the laboratory are not considered sufficiently accurate for the space mission: The radiation sources do not fully reproduce SSI. Additionally, through integration into the spacecraft, their qualifications with thermal cycling and vibrational tests, transportation to the launch, and the launch itself, the calibration parameters undergo changes from the moment they are set in the laboratory up until first being applied in space [17]. The instruments also remain under significantly changing environmental conditions for several months without any chance of recalibration.

B. Calibration in Space

During the commissioning phase, the laboratory procedures and the expected SSI signals must be checked and adapted to conditions in space.

After testing the functional parameters in terms of housekeeping data and spectral scans with and without filters, DARA, IC, PMT, and BOS recordings will help to update the laboratory efficiency curves, allowing SSI data to be derived directly from the SP recordings.

In order to update the smooth efficiency curves ${\eta }_{\mathrm{SPACE}}$, three to five filters are chosen for each SP, depending on the wavelength range (see Section 3.A),

The filters are not required to completely cover the spectral ranges because interpolation from one spectral range of the filters to the other is sufficient due to the smoothness of the efficiency curves.

Depending on the SNR for different spectral regions, the sequences and repetition rates will be optimized such that filter intervals of lower SSI throughput will receive higher measuring repetition rates.

C. SOLACER: Internal Cross-Calibration

Internal cross-calibration is one of the keys to achieving high data accuracy. There are quite a few possibilities.

It is of utmost importance to cross-calibrate the secondary irradiance detectors to the DARAs and ICs in order to take advantage of the BOS and PMT detectors’ increased sensitivity. This provides substantially higher accuracy in the spectral regions where SSI energy is too low for the DARAs to generate useful signals. We call this procedure “BOS and PMT detector bridging from DARA to IC technologies” between about 120 and 250 nm of SSI.

With reference to Table 1, there are three spectral ranges overlapping with the neighboring SPs at both the outer and inner ring of the filter wheel. So, a further task is to choose three filters that will cover each of these overlaps for cross-calibration purposes.

Different gases with overlapping ranges of ionization (Fig. 4) add to the internal cross-calibration capabilities, among other things.

Both DARAs are also operated in the TSI mode, enabling cross-checking with data from other TSI missions and comparison with integral SSI data.

D. Operations in Orbit

The possibility of collecting data at eight SPs, two DARAs, and two BOSs simultaneously offers the chance to record a great number of SSI spectra and provides long periods of time for calibration, which is useful for allocating longer measuring periods for spectral intervals with lower SNR.

The commissioning phase will also be used to check the statistics of the spectral DARA and BOS recordings in order to expand on the measuring sequences, which is important for optimizing how the four parameters of filter bandwidth, the period and repetition of measurements, aperture size, and detector sensitivity interact to determine the accuracy of the SSI data. One example is elaborated in Table 4. For absolute calibration, one IC measurement per day (or even per week) is sufficient at the beginning of the mission. After the first six months, execution of the calibration mode can be reduced to every second week. The same is valid for the DARA measurements. However, selected intervals such as that of 150–250 nm are calibrated more frequently than others.

Table 4. SOLACER Expected Accuracy with DARA, IC, BOS, and PMT Detectors

View Table | View all tables in this article

In general, calibration will be achieved on demand.

To evaluate and compare data recorded at different times, a solar EUV activity sensor (SEPS) is operated continuously with a temporal resolution of 0.1 s (see Section 4.F).

E. Expected Accuracy Achievement

SOLACER SSI SPs operate at relatively high levels of data statistics [11,23], providing high-precision data at high SNRs (see also Table 2).

The realization of the SOLACER instrumentation shows great promise for substantially improving the insertion of TSI detectors and ICs that remain stable in the long term:

State-of-the-art TSI radiometers have accomplished an absolute accuracy of $0.3\mathrm{W}{\mathrm{m}}^{-2}$ or $2.2*{10}^{-4}$ [20,21], and ICs reach an absolute accuracy of $\sim 5\%$ regarding expected accuracy levels, as summarized in Table 4.

A TSI DARA detector generates a minimal useful signal of $S_{\text{SSImin}}(r=5\mathrm{mm})=0.5\mathrm{W}{\mathrm{m}}^{-2}$ with a circular aperture of 5 mm diameter of $A_5=19.6{\mathrm{mm}}^{-2}$. Implementing an aperture of 8 mm diameter with $A_8=50{\mathrm{mm}}^2$, the minimal useful signal is

To achieve calibrated SSI data with a statistical noise $<1\%$, the filter bandwidths $\mathrm{\Delta }{\lambda }_F$ and the number of repeated measurements $N$ are the parameters to be optimized. For wavelengths $>290\mathrm{nm}$, each signal of DARA shows an accuracy $<1\%$. Below 290 nm, the measurements are to be repeated. Supposing satellite orbits of 90 min with a sunlit period of 45 min, $N=2*45=90$ accumulated data points (fifth column, mid-subcolumn) are recorded providing signals

The accuracy of filter transmission is significantly $<1\%$, based on the $\mathrm{SNR}>{10}^4$. For the circular size of the aperture with $\pm 10\mathrm{\mu m}$ uncertainty, the accuracy is $\ll 1\%$.

The higher sensitivity of the internally cross-calibrated BOSs allows a direct bridging from 300 to 126 nm where the ICs are operated. This way, bridging from the VIS to the EUV spectral regions is a central point of SOLACER measuring technique to improve the accuracy of the SSI in the VUV spectral range.

Solar modeling delivers another important tool to increase accuracy, especially of the bridged VUV spectral range, relating the different spectral ranges to each other over the TSI.

Having in mind that up to 16 orbits per day will be available, there is a lot of time to recalibrate SPs every month–a comfortable possibility in view of the slowly changing efficiency parameters during long-term missions. Strongest efficiency changes occur during the launch and first weeks in operation.

Absolute calibration checks shall also be done comparing the sum of the spectral SSI data to TSI. Though this is quite a complex procedure, it shall be developed for stepwise approaches.

Since planar grating SPs of SOLACER would become too large, providing very high spectral resolution, and since traditional SSI SPs (Section 2.A) do not meet the requirements for absolute calibration, a combination of SOLACER and a high-resolution SP would be a good choice for covering the full SSI spectral ranges.

F. Special Features to Increase SOLACER Capabilities

If a planar EUV and plasma sensor of 6 cm diameter and about 150 g recording the integral EUV fluxes at 0.1 s temporal resolution [17] will be inserted, solar activity can be observed, adding to the data evaluation, for example, to recognize and correct for short-term SSI fluctuations.

This EUV and plasma sensor and the SPs will also record proton events and background signals from the Atlantic Anomaly [31], providing measures to correct SSI data, if required.

In the IR spectral region, correction of the thermal background emitted from the filters during the DARA and BOS measurements can be determined with obstacles in front of the filters, stopping the SSI heating up the filters while the obstacles will be moved back and forth at different periods during test orbits. The change from the sunny to the shadow sides will also provide information on the thermal contribution of the filters to the DARA and BOS signals.

Wavelength determination will be achieved by small laser diodes, narrowband LEDs, irradiating mechanical holes, hollow cathode lamps operated with idle gases such as argon, and other devices.

5. SUMMARY AND CONCLUSIONS

From the systematic reanalysis of the requirements for future SSI observations, it is concluded that neither laboratory precalibration nor standard radiation sources in space can accurately determine degradation of SSI SPs. Correcting the data for degradation with the natural radiation source of the Sun is an important requirement to accomplish SSI data of high accuracy, leading to the expansion of the SolACES SP system to measure the SSI radiation from the XUV to the IR spectral regions and to repeatedly recalibrate the SPs. The SolACES instrument has been operating for nine years aboard the ISS while extremely strong degradation was experienced. It was corrected by the experimental resources with ICs aboard. Based on the same principle of degradation correction, a new instrument, SOLACER, has been designed to cover the (2)-17-2800 nm spectral range by adding two radiometers (DARA), two BOSs, and two PMTs. Bridging the DARA and the IC spectral ranges of sufficient strength of their signals and applying internally cross-calibration BOS and PMT techniques are key to the performance of SOLACER. It is expected that this new instrument for SSI and TSI measurements will provide data with an unprecedented accuracy also for 11-year missions and beyond. SOLACER is considered the first autocalibrating SSI system based on primary irradiance detectors. Given the state-of-the art accuracy of $2*{10}^{-4}$ for the TSI detectors and $5*{10}^{-2}$ for the ICs, the accuracy numbers listed in Table 4 shall be improved further for next-generation instruments to approach ${10}^{-3}$ accuracy. Also, $5*{10}^{-2}$ for the ICs is not a physical limitation of IC accuracy.

Taking into account this future outlook, there is a good potential for SOLACER serving as a reference for other SSI SPs in space.

In the fields of solar, solar-terrestrial, atmospheric, and climate physics, an optimal payload for SSI measurements will comprise SOLACER, a high-resolution SP to be cross-calibrated by SOLACER and a TSI instrument.

Because TSI and BOS detectors are also well suited to observe the Earth IR and the reflected solar radiation, a combined payload of solar-oriented TSI and SSI instruments and terrestrial-oriented TSI and BOS sensors provides a good opportunity to acquire high-quality data sets for climate modeling with respect to the solar-terrestrial energy balance–a compact payload $<70\mathrm{kg}$ well suited aboard a platform like the Airbus DS Astrobus S line.