Content

Remote Sensing from Space

You are here: Remote Sensing from Space Projects Diarad/VIRGO The instrument

The DIARAD/VIRGO instrument performance

1. Short description and design particularities of DIARAD
2. Performance of DIARAD in space
3. Electrical characteristics
4. DIARAD and solar noise measurements
5. Temperature sensivity of DIARAD
6. DIARAD pointing sensivity
7. DIARAD data product
7.1 At level 1 DIARAD provides only one product
7.2 Data treatment from level 0 to 1
8. DIARAD quality assessment
9. DIARAD solar constant time series
9.1 Short history of operations
9.2 DIARAD radiometric aging
9.3 The solar constant

1. Short description and design particularities of DIARAD

DIARAD is a Differential Absolute Radiometer. It is composed of two cylindrical cavities coated inside with diffuse black and mounted next to each other on the same heat sink. The flat bottom of the cavities are in fact heat flux transducers on which heating elements have been mounted. Both cavities see the same thermal environment through accurately know circular apertures. The principle of operation is very simple : it is based on the comparison of the power generate inside the cavities very similarly as an househould weight can be used. For instance a constant electrical power is generated in one of the channels and the difference between the two heatflux sensors is automatically brought back to zero by and ad hoc accurate servosystem that provides electrical power to the other channel called "active channel". This one is regularly irradiated by the Sun or closed. The difference of the electrical power fed to the active channel when its shutter is open (exposed to the Sun) and when it is closed is proportionnal to the incident solar irradiance. The factor of proportionnality is a function depending on the following elements determined accurately in our radiometric characterisation laboratory : the area of the circular aperture in front of the cavities, the geometrical absorption coefficient of the cavities (function of the absorbing paint), the (thermal) efficiency factors of the cavities, the shutter thermal emission, the diffraction, as well as a number of small corrections keeping account of parasitic scattering and other secondary imperfections [1]. Both channels have been characterised independently as well for air as vacuum conditions. All of our absolute radiometers, all of the same design principle (SOLCON, SOVA 1 & DIARAD) are radiometrically independent of each other as well as of the World Radiometric Reference (WRR). Of course whenever we could, we have compared them to each other as well as to the WRR. In particular, SOLCON and SOVA 1 left and right channels bear the heritage of the Space Absolute Radiometric Reference (SARR) determined during the ATLAS 2 mission. 

2. Performance of DIARAD in space

Measured performances of DIARAD in space refers only to its electrical performance characteristics spread over a period of about four months for which DIARAD was almost all the time operated with its left channel active, providing every three minutes a solar irradiance measurement. At selected times, seven in total, the same operation was performed for a duration of half an hour on the right channel preceded by a long sequence of a couple of hours of different radiometric states for calibration and health monitoring.

3. Electrical characteristics

The DIARAD left and right channels electrical power are measured with four differential amplifiers in sequence quasi simultaneously. Their output is fed to the VIRGO data acquisition system (DAS) which converts their output voltages to counts that are transmitted to the Earth. The knowledge of their gains and offsets behaviour is thus critical to the long run accuracy of the observed solar irradiance. For this reason, the input of the four amplifiers is switched every two to three days to the output of a highly stabilised and accurate voltage reference built into DIARAD. The output of the amplifiers have been measured by the VIRGO data acquisition system (DAS) with a stability over four months of the order of 10-5 (two high accuracy voltage references). The inflight measured gains and offsets of the four DIARAD amplifiers showed some noise and trends but are well described each by one number within the margin given in the following table. Channel Gain (G) deltaG Offset (O) deltaO Voltage left 0.914868 ± 8 10-6 1.43 10-4 ± 1.12 10-4 Voltage right 0.914960 ± 1.8 10-5 4.26 10-4 ± 1.0 10-4 Current left 1.306970 ± 1.0 10-5 - 9.83 10-5 ± 7.0 10-4 Current right 1306092 ± 1.5 10-5 2.66 10-4 ± 1.0 10-4 Taking into account the values of the uncertainties of the resistors used to measure the currents as well as the input voltages to the four measurement amplifiers, one finds that the incertitude margin on the voltages and currents is within the range of ± 3.0 10-5 during the first four months of operation of DIARAD with Sun observations. The gain and offset monitoring of DIARAD will be monitored during the full life of DIARAD. It will serve to guarantee its electrical accuracy as well as to monitor and access the stability of the PMO radiometers whose measurement amplifiers gains and offset have been determined only on the ground before launch. The expected accuracy incertitude over the four months on the determination of the electrical power is thus within a relative range of ± 6 10-5 . This is excellent considering that the measurements are made in space. On the ground we usually verify the necessary (not sufficient) condition that the heating resistors comply with the values measured independently, to have an idea of the achievable power measurement. Over six months we find the right heater value to be within a relative margin of ± 4 10-5 and for the left heater to be within ± 10-4. Compared to the nominal values measured one year before delivery the relative margins are ± 10 -4 for the right heater and ± 3 10-4 for the left heater. The wider margin than for the expected power measurements is coherent with heater aging effects that is easily acceptable at the left side due to the quasi continuous open-close cycling and solar exposure of that radiometric channel.

4. DIARAD and solar noise measurements

In view to be able to objectively identify solar noise from DIARAD instrumental noise we have made the histograms of three times two days during the first four months solar observations. The mean of their characteristics are shown in the following table. Left Left Right channel channel channel close open close Standard 2.5 10-3 6.2 10-3 8.7 10-4 deviation (mW) Mean (mW) 107.48 39.55 104.01 Relative 2.4 10-5 1.6 10-4 8 10-6 Standard deviation Kurtosis 0.016 - 0.59 + 2.56 The kurtosis shows, as expected, that the right reference channel is not subject to a random distribution like the left channel which for the closed case is very nearly random (effect of the servo system hunting the right reference + a constant system offset). When the left channel is open, the signal deviates from random, the histogram being flatter than for a normal distribution, together with a relative standard deviation about seven times higher than the relative standard deviation when the left channel is closed, this is the signature of the solar noise. We suppose that the fluctuations of the signal with the left channel open is caused by the combination of stockastically independant instrument noise and solar noise, and that the instrument noise standard deviation is the same with the channel open or closed. Therefore, the relative standard deviation of the solar constant is about 8.3 10-5 which corresponds to the noise range observed. The measurement resolution is finally limited by the noise of the DAS.

5. Temperature sensivity of DIARAD

The maximum temperature variation up to end of July 1996 has never been higher than 2.5 C, in particular with a jump of 1.5 C that occurred for a duration of about 18 days. Otherwise a slow upward trend of about 1 C over the four months can be identified. The DIARAD radiometer is particularly immune to temperature effects, this is due to its differential design up to now unique within the set of radiometers that are used to observe the solar constant. Of course a temperature coefficient is applied in the DIARAD algorithm to take care of the sensitive area value, but no valid temperature coefficients could be determined yet for the gains and offsets of the four measurement amplifiers. The shutter effect correction is based on the direct measurement of its temperature.

6. DIARAD pointing sensivity

The classical pointing correction is foreseen for DIARAD, due to the accuracy of the SOHO solar pointing. This correction has not been applied up to now.

7. DIARAD data product
7.1 At level 1 DIARAD provides only one product

the solar constant measured with the left channel of its differential absolute radiometer every three minutes. Continuous health analysis are performed based on the continuity of its measurable characteristics like the values of the heaters, the gains and offsets of the measurement amplifiers, the noise levels, the instrument temperatures. In addition wavelets analysis are run and accumulated to monitor the non stationary temporal behaviour of the solar constant and its noise spectra [2]. We hope that by doing this we accumulate data useful for the finding of the g modes of solar oscillations.

7.2 Data treatment from level 0 to 1

The data treatment applied to the counts produced by DIARAD is described in a definition document [3]. It consists essentially to convert the counts to voltages and currents, they are themselves converted to resistance values and electrical power. The solar constant is then a function of electrical power difference between the open and closed state of the active radiometric channel to which is applied a number of factors to take account of the radiometric characteristics (sensitive area, efficiency factor, diffraction effect, shutter effect, temporal dissymetry of reference channel, etc...) the SOHO to Sun distance and if necessary the pointing offset. In principle, the level 2 data of DIARAD will be very similar or identical to the level one. The complete data treatment will be revisited, the reference document cleaned up and hidden approximations will be chased. A second product will be added to the original solar constant measured by DIARAD, this is the SARR based solar constant derived from the SOLCON HITCHHIKER mission.

8. DIARAD quality assessment

The VIRGO flight conditions are completely different from the conditions experienced on board of a polar platform like EURECA or the shuttle on board of which flew the radiometers SOLCON and SOVA 1 based on the same basic radiometric differential design as DIARAD. Because of the smaller sensitive aperture and the different electrical design constraints, the relative error of DIARAD is tought to be higher than SOVA 1 and SOLCON. However as the expected assymptotic minimum of the solar constant during cycle 21 was 1365.02 Wm-2 and as the mean observed with the left channel of DIARAD is 1366.00 Wm-2, this channel fits probably with the set of left and right channels of SOLCON and SOVA 1, all within ± 0.1 % dispersion. The right channel reads about 5.87 Wm-2 higher. This is unusual compared to the set of instrument that have been constructed at the Royal Meteorological Institute up to now. Except of the fact that it is the first time that some superstructure is put above the shutters of DIARAD as well as the presence of a flushing tube near by or a some wire providing diffused light parasitic reflections, we could not find any coherent explanation of such an high reading with the right channel of DIARAD. For sure this mystery will never be resolved. The right channel will thus only be used to monitor the aging of the left channel. Most probably mid 1997, SOLCON will be launched on the space shuttle as a HITCHHIKER payload. This will allow to localise DIARAD with respect to the Space Absolute Radiometric Reference (SARR) defined during the ATLAS 2 mission, when ten radiometers have measured the solar constant during a particularly quite solar activity period. SOLCON as well as SOVA 1 bears the heritage of the ATLAS 2 mission [4]. By doing this, the solar constant observations made from VIRGO will be provided with an ad hoc adjustment coefficient to insure the continuity at climate scale of the SARR adjusted observations made since 1978.

9. DIARAD solar constant time series

9.1 Short history of operations

Since the fourth day of the mission, DIARAD is powered on continuously. Except for one day, a set of different radiometric modes has repeatedly been run for electrical calibration and health monitoring up to the continuous solar irradiance observation period performed with the left channel in the servoloop and its shutter opening and closing every 90 seconds and the right channel serving as the reference with its shutter always closed. From time to time, the roles of the left channel and the right channel are reversed for half an hour with the purpose of monitoring the aging of the continuously exposed left channel.

9.2 DIARAD radiometric aging

The following table provides the results of the left channel aging monitoring. One must remember that the dispersion of the difference between the solar irradiance measured with the left and right channel is essentially due to the solar noise. Date of comparison 26/01 07/02 07/03 06/04 04/06 (1996) Right - left 5.84 5.93 5.83 5.88 5.88 (difference Wm-2) Cumulate left 9120 24702 65424 107280 190443 channel - Solar exposure (minutes) Cumulate right 60 90 120 150 180 channel - Solar exposure (minutes) From the observed differences over a duration of four months it is not possible to deduce objectively an aging of the left radiometric channel, a longer time period seems to be necessary. Compared to the aging of the SOVA 1 left channel at EURECA altitude of 500 Km a none linear aging of a total of 580 mW/m-2 in 300 days was observed. The situation is quite different. The DIARAD behaviour with its cyclindrical cavity covered with diffuse black seems to be very encouraging, perhaps this is due to be absence of atomic oxygen compared to its density at the EURECA altitude.

9.3 The solar constant

The actual observation period corresponds to very low to low solar activity. The lowest montly smoothed sunspot number is predicted to be in November-December 1996 (SIDC). Our solar constant observation do not show yet any increasing trend. Only a few bumps (faculae) reappearing after one solar rotation together with a small dip can be observed.