- Email: [email protected]
A microsatellite platform for hot spot dedection
Acta Astronautica 56 (2005) 221 – 229 www.elsevier.com/locate/actaastro
A microsatellite platform for hot spot dedection Ingo Walter∗ , Klaus Briess, Wolfgang Baerwald, Eckehard Lorenz, Wolfgang Skrbek, Friedrich Schrandt DLR, Institute of Space Sensor Technology and Planetary Exploration, Rutherfordstr. 2, 12489 Berlin, Germany
Abstract The main payload of the BIRD micro-satellite is the newly developed hot spot recognition system. Its a dual-channel instrument for middle and thermal infrared imagery based on cooled MCT line detectors. The miniaturisation by integrated detector/cooler assemblies provides a highly efficient design. Since the launch in October 2001 from SHAR/India the BIRD payload, claiming 30% of the BIRD mass of 92 kg, is fully operational. Among others forest fires (Australia), volcanoes (Etna, Chile) and burning coal mines (China) have been detected and their parameters like size, temperature and energy release could be determined. As the status of the payload system is satisfactorily it has a potential to be applied in new missions with the help of modern detector technology. © 2004 Elsevier Ltd. All rights reserved.
1. Introduction
oriented on a set of ambitious objectives:
There are several operational systems used for global observations of active fires (e.g. NOAA, EOS, ENVISAT), nevertheless the technology to be operated comprises remarkable potential of further developments, for instance, aspects of resolution and saturation of the sensors. Based on the core activity to design a new generation of imaging infrared detectors the interest to investigate their operational behaviour in space led to a complete micro-satellite development, which is
1. Development, qualification and demonstration of small satellite technologies in space. 2. Test of a new generation of infrared array sensors adapted to Earth remote sensing objectives by means of small satellites. 3. Detection and scientific investigation of hot spots (forest fires, volcanic activities, burning oil wells or coal seams).
∗ Corresponding author. Tel.: +49 30 67055 186;
fax: +49 30 67055 532. E-mail address: [email protected] (I. Walter). 0094-5765/$ - see front matter © 2004 Elsevier Ltd. All rights reserved. doi:10.1016/j.actaastro.2004.09.009
2. Scientific instruments of BIRD The temperature of vegetation fires varies from 500 up to 1200 K depending on the type of combustion. Their spectral density distribution occurs in the midwave infrared (MIR) wavelength region at 3–5 m.
222
I. Walter et al. / Acta Astronautica 56 (2005) 221 – 229
Table 1 BIRD multi-sensor system parameters (altitude 572 km)
Wavelength Focal length Field of View F-number Detector Detector cooling Pixel size Number of pixels Quantization Ground pixel sizea Swath widtha
WAOSS-B
MIR
TIR
600–670 nm 840–900 nm 21.65 mm 50◦ 2.8 3 CCD-line arrays Passive, 20 ◦ C 7 × 7 m 2 2884 11 bit 185 m 533 km
3.4–4.2 m
8.5–9.3 m
46.39 mm 19◦ 2.0 MCT-line array Stirling, 105 K 30 × 30 m2 2 × 512 staggered 14 bit 370 m 190 km
46.39 mm 19◦ 2.0 MCT-line array Stirling, 80 K 30 × 30 m2 2 × 512 staggered 14 bit 370 m 190 km
WAOSS-B—Wide angle optoelectronic stereo scanner, modified for BIRD, MIR—Medium infrared sensor, TIR—Thermal infrared sensor. a Orbit height = 572 km.
Hence it is the main target for a space-born fire sensor. Reflection effects (e.g. sun glint) as sources of data misinterpretation have to be distinguished by combination of MIR and VIS (visible) channels. Furthermore, temperature information has to be derived from thermal infrared (TIR) imaging data by bi-spectral data processing techniques. The so arising multi-channel system is able to detect hot spots of sub-pixel size.
2.1. Payload architecture and parameters The payload is designed to fulfil the scientific requirements under the conditions of a microsatellite. It consists of the following main parts (Table 1): • Bi-spectral infrared HSRS (HSRS including MIR and TIR-Sensor), • Wide-angle opto-electronic stereo scanner (WAOSSB, VNIR-Sensor) for vegetation analysis and fire classification, • Payload data handling system (PDH) with a mass memory, • A neural network classificator for an on-board classification experiment.
2.2. HSRS—the main BIRD payload The main payload of the BIRD micro-satellite is the newly developed hot spot recognition system (HSRS). Its a dual-channel instrument for middle and thermal infrared imagery based on cooled MCT-line detectors. These linear arrays consisting of 2 × 512 pixels in staggered configuration are based on loophole diode technology with detector electronics on a hybridised detector assembly. Their operating temperatures are 80 K for the TIR and 105 K for the MIR channel, respectively. Based on experiences with a airborne laboratory model using a split Stirling cooler design the detector assembly was optimised regarding stray-light suppression and power needs to meet the radiometric requirements and to cope with the limited resources of the micro-satellite concept. A detector assembly for military applications (see Fig. 1) was modified into a space-qualified solution. The main aspects of the sensor design are: • integrated miniaturised Stirling cycle cooling device, • adapted cold shield design inside the Dewar covered by the IR-window,
I. Walter et al. / Acta Astronautica 56 (2005) 221 – 229
223
Stirling Cooler Motor incl. Drive electronics
Cooled Finger Dewar Dewar Flange incl. Electr. Connectors
Cold Shield Assembly Crank Mechanism Body IRwindow
Focal Plate Assembly incl. MCT Detector
MCT array 2x512
Fig. 1. Miniaturised integrated detector/cooler assembly.
• •
30 x 30 µm pitch, loop-hole diodes
Fig. 2. Design overview of the HSRS detector assemblies.
• special cold finger interfaces for stiffness against dynamical loads with sufficient thermal decoupling. This system consumes 11 W, which is a factor of 4 less in comparison to a split cooler concept. Fig. 2 provides the main constructive aspects of the integrated detector cooler assembly. The design of the flight configuration of the HSRS is driven by strong requirements of the coalignment stability in the line-of-sight of the MIR to the TIR channel (±0.2 arcmin at duty cycle). It is characterised by the following attributes: • identical construction of the optical channels (only optical coating of lenses different, Fig. 5), • autonomous thermal control including heat storage and radiator, • opto-mechanical CFRM-structure with high resonance (> 200 Hz). Fig. 3 provides the overview of the MIR/TIR sensor head design and Fig. 4 gives an impression of the flight hardware completed by thermal control components like radiators and multi-layer insulation. A serious problem of the push-broom sensor, especially in the TIR region is the radiometric stability of the detector elements during the data records which
cannot be covered by laboratory ground calibration. Hence is it essential to add an in-flight calibration device. Thus the HSRS has black body units as temperature reference targets in front of the IR-lenses which can be exposed to the complete field of view at the beginning and the end of the data records. In the passive periods this device serves as protection cover of the optics against contamination and undesirable cool down. The inherent total failure risk is minimised by a separation mechanism that rejects the cover system in the case of emergency. 2.3. WAOSS-B—an example of hardware re-use The WAOSS-B is the VIS/NIR-sensor of the BIRD payload. Actually developed for the MARS-96 mission a flight spare model was modified slightly for the application on BIRD (Fig. 6). A special lens design with integrated filter provides narrow spectral bands over a wide field of view of 50◦ (see Fig. 6). It adopts the instrument to the tasks of vegetation exploration as presented in Section 4. The in-track stereo capability is an important feature for the cloud investigations.
224
I. Walter et al. / Acta Astronautica 56 (2005) 221 – 229
Stirling Cooler Engine Optics Block (Invar) Focusing / Alignment Device
IR-Baffle incl. Cover IF
Optomechanical Structure incl. Heat Storage Detector Proximity Electronics
Cover Separation Device
IR-Dual Band Lenses Inflight Calibration Unit incl. Black Body
Cover Drive Mechanism
Fig. 3. HSRS sensor head design overview.
3. Multi-sensor assembly on a microsatellite 3.1. Structural concept of BIRD
Fig. 4. HSRS with IR electronics unit in flight configuration.
Due to the modular electronic concept of WAOSS-B almost no hardware modification was needed and the reconfiguration is concentrated on the software side, i.e. to provide the master clock function for simultaneous imaging of all BIRD instruments (Fig. 7).
Classical spacecraft designs are oriented on a clear separation of payload which are the scientific instruments and the spacecraft bus containing all subsystem supporting the function of the payload. Here a strong interaction during the design process requires strict management of interfaces from a very early design phase on. In configurations of small satellites this in addition takes place within a small volume. Therefore for BIRD payload and subsystem components had to be developed in a highly iterative process considering the satellite as one complex device. Interface modifications were controlled by defining a modular system for the structure. Thus the BIRD architecture is a cube-shaped tower of three segments (Fig. 8): • service segment as the basic segment with separable launcher interface containing batteries, reaction wheels and gyroscope as well as the GPS system,
I. Walter et al. / Acta Astronautica 56 (2005) 221 – 229
225
Fig. 5. Preintegrated optical bench with the dual lenses and CFRM-structure for thermo-optical stability.
Payload Segment
S-Band Antenna
Electronics Segment Service Segment
Fig. 8. Flight configuration of BIRD (0.6 × 0.6 × 1.6 m3 ).
Fig. 6. Spectral bands of WAOSS-B (NIR doubled).
• electronics segment as a container of payload, spacecraft electronics and the communications package, • payload segment as a platform for the scientific instruments and the star sensors. The configuration is completed by a solar cell system of three panels—two of them deployable—the antennas and magnetic torquers. 3.2. Payload platform concept and technology
Fig. 7. WAOSS-B under integration (8.4 kg, 0.18×0.21×0.38 m3 ).
Due to the power needs of the payload of 200 W in full operation the BIRD mission concept is based on intermediate operation of the scientific instrumentation of approximately 30 min per orbit (10 min data take) and the sun pointing for most effective energy reconstitution. Here the challenge is to cope with the cyclic thermal loads and their influences on the geometric-optical stability of the instruments
226
I. Walter et al. / Acta Astronautica 56 (2005) 221 – 229
Fig. 9. Design principle of the payload platform (cross section).
Fig. 11. BIRD ready for launch as a piggy-back payload on the Indian PSLV C3 vehicle.
Fig. 10. CFRM platform incl. payload mounting points and thermal control elements.
nections with thermal relevance are glued with silverfilled epoxy. The heat conduction reaches over 50% of that from aluminium. Its CTE instead is a factor of 15 less. Another effect is the decoupling of structural loads coming from the spacecraft bus because of separated mounting planes for spacecraft and scientific instruments.
4. In-orbit results line-of-sights which have to be stable within a subpixel range. Early considerations showed the necessity of a thermal control of 1 K for the instrument structure if a conventional isostatic aluminium plate would be used. Despite the large mass of such a solution that was not feasible within the microsatellites resources. The resulting technical solution of the instrument platform is to combine two CF-honey-comb panels on top and below of a 3-mm-thick layer of carbon fibre carbon (CFC). This layer provides heat conduction to the heat pipe interfaces on the front ends of the platform with a coefficient of 155 W m−1 K −1 —comparable with aluminium. The instrument mounting interfaces of this multisandwich structure are inserts with an enlarged plate founded in the CFC-layer (see Figs. 9 and 10). All con-
With the launch on 22 October 2001 BIRD (Fig. 11) was set into a circular sun-synchronous orbit with an altitude of 572 km. Ever since several scientific results regarding hot spot hot detection have been achieved like: • wild fire detection in Brazil (simultaneously with other space-born sensors), • bush fire monitoring in Sydney and Canberra, • investigation of volcanic activities in Chile, Mexico and Hawai, • determination of industrial hot spots in Europe, • coal seem fire investigation in China. The calibration status, especially of the HSRS is monitored by comparisons of temperature data over water
I. Walter et al. / Acta Astronautica 56 (2005) 221 – 229
227
Fig. 12. Vegetation re-growth after fire in New South Wales (Australia, January 2003 [1]): (1) geographical map; (2) MIR image with fire fronts; (3) NIR image with burned areas; (4) NIR image 2 months later.
and during each duty cycle via channelwise imaging of the black bodies. Also a ground-truth measurement campaign in Bavaria in January 2003 was successful. A 2 × 2 m2 sized bonfire (900 K) was lit and measured in parallel
to the BIRD imaging proving the high sensibility of the system beyond its nominal lifetime of 1 year in orbit. Apart from detecting hotspots the images of WAOSS and HSRS can be used to determine the area
228
I. Walter et al. / Acta Astronautica 56 (2005) 221 – 229
affected by the fire allowing e.g. an estimation of the extent and consequences of vegetation fires. If the area is surveyed in regular intervals the vegetation re-growth can be monitored. An example is given in Fig. 12. 4.1. Pixel coregistration aspects The coregistration of the pixels is essential to acquire the accurate values for temperature and area of the hot spots as well as their energy release. Even for hot spots in the sub-pixel range (typ. size down to 12 m2 , T = 1000 K) this is important. Based on the data ground processing which is performed to generate the hot spot characteristics it is possible to derive status information about the payload configuration in space. Table 2 lists typical values, derived from [2] of the pixel displacement of the row data which have to be corrected during the matching process. Here Y is the flight direction while X is the across-track direction. Although tilt effects and time-related influences especially in flight direction cannot be assessed it can be seen clearly that the NIR cannel shows remarkable more dynamics than the TIR channel. Projected to the payload segment design that means that the WAOSS-B as a stand- alone instrument is influenced more than the IR-channels which are
designed as twins on a common opto-mechanical structure. Here the order of magnitude is the same as found in the vacuum in situ tests. A new quality of the BIRD data is its geo-referenced format by means of the correction of all flight attitude data. Here also a measure is given for the coalignment status. However in this case the complete error chain of deviations is covered. This work is under progress but 200 m misregistration typical for the WAOSS-B instrument is reported in [1].
5. Growth potential Although the BIRD payload was developed in a straightforward manner typical for micro-satellites since the origin of its IR-detector technology almost a decade has past. So it is worth to review the latest developments to prove their potential for new applications. Here the recently upcoming combination of two spectral bands sensitivities on a common pixel in the infrared range is of interest. This dualband detectors take advantage of the advances in micro-structures fabrication (see Fig. 13) providing photosensitive areas in the MIR and in the TIR
Table 2 Record of the misalignment correction for ground processing of BIRD data over 8 months Image date
011123 011218 020208 020215 020320 020323 020326 020327 020409 020471 020418 025017 020718 020719
Image target
Europe US China China China China Benin Titikaka Reunion Australia Australia Borneo California Etna
Channel displacement related to the MIR image in pixels NIR (Y, X)
TIR (Y, X)
−4, −1, −5, −4, −5, −5, −7, −2, 0, 0, −4, −6, −4, −5,
0, 0, −1, −1, −1, −1, −1, −1, −1, −1, −1, −1, −1, −1,
2 7 5 4 −9 4 1 6 5 5 6 −2 −10 6
0 0 0 0 0 0 0 −1 −1 −2 −2 −1 −1 −1
Fig. 13. Micro-structured epitaxial layers for 2-colour MCT detector (from [3]).
I. Walter et al. / Acta Astronautica 56 (2005) 221 – 229
229
Fig. 14. BIRD payload platform (460 × 460 mm2 ) with the potential of higher integration by application of modern technology.
Fig. 15. Miniaturised interferometer principle [4]: (1) retro-reflectors, (2) pendulum system, (3) beamsplitter, (4) detector system.
within a pixel pitch. This was demonstrated for quantum-well detectors already and now extended to MCT detectors too. Besides the integration of the spectral bands the sandwich-type configuration provides an excellent coregistration of the pixels. It is by far the best approach to meet the related requirements for the hot spot detection. Projected to the BIRD payload platform this gives the opportunity to integrate the channels of the HSRS into one common opto-electronical system while the exempted part is filled by a modern VIS/NIR-imaging
system. Hence the HSRS would have integrated the functions of WAOSS with the result of having onethird of the payload platform space empty for new instrumentation. This is illustrated in Fig. 14. One idea for that is to add a spot sounding spectrometer for the assessment of gases and aerosols for an enhanced hot spot detection technique. A flight proven design according [4] is a double-pendulum interferometer which provides a robust design with maximised compactness reaching a high optical throughput (see Fig. 15). The described approach would keep the hot spot detection system suitable for small satellites with substantial improvements of its performance. It is at least a contribution to a cost effective and consequently realistic space-borne disaster monitoring system. References [1] E. Lorenz, et al., Scientific and operational exploitation of BIRD, FIREBIRD Technical Notes, Report to ESA, 2003, p. 22 ff. [2] B. Zhukov, et al., BIRD detection and analysis of hightemperature events: first results, Proceedings of SPIE 4886, 2002. [3] W. Ziegler, et al., 3. Generation developments in FPAtechnologies, Company Presentation AIM GmbH, 2003, p. 3. [4] H. Hirsch, Optical design and performance of the planetary fourier spectrometer (PFS), Mikrochim Acta (Suppl.) 14 (1997) 571–574.
A microsatellite platform for hot spot dedection Ingo Walter∗ , Klaus Briess, Wolfgang Baerwald, Eckehard Lorenz, Wolfgang Skrbek, Friedrich Schrandt DLR, Institute of Space Sensor Technology and Planetary Exploration, Rutherfordstr. 2, 12489 Berlin, Germany
Abstract The main payload of the BIRD micro-satellite is the newly developed hot spot recognition system. Its a dual-channel instrument for middle and thermal infrared imagery based on cooled MCT line detectors. The miniaturisation by integrated detector/cooler assemblies provides a highly efficient design. Since the launch in October 2001 from SHAR/India the BIRD payload, claiming 30% of the BIRD mass of 92 kg, is fully operational. Among others forest fires (Australia), volcanoes (Etna, Chile) and burning coal mines (China) have been detected and their parameters like size, temperature and energy release could be determined. As the status of the payload system is satisfactorily it has a potential to be applied in new missions with the help of modern detector technology. © 2004 Elsevier Ltd. All rights reserved.
1. Introduction
oriented on a set of ambitious objectives:
There are several operational systems used for global observations of active fires (e.g. NOAA, EOS, ENVISAT), nevertheless the technology to be operated comprises remarkable potential of further developments, for instance, aspects of resolution and saturation of the sensors. Based on the core activity to design a new generation of imaging infrared detectors the interest to investigate their operational behaviour in space led to a complete micro-satellite development, which is
1. Development, qualification and demonstration of small satellite technologies in space. 2. Test of a new generation of infrared array sensors adapted to Earth remote sensing objectives by means of small satellites. 3. Detection and scientific investigation of hot spots (forest fires, volcanic activities, burning oil wells or coal seams).
∗ Corresponding author. Tel.: +49 30 67055 186;
fax: +49 30 67055 532. E-mail address: [email protected] (I. Walter). 0094-5765/$ - see front matter © 2004 Elsevier Ltd. All rights reserved. doi:10.1016/j.actaastro.2004.09.009
2. Scientific instruments of BIRD The temperature of vegetation fires varies from 500 up to 1200 K depending on the type of combustion. Their spectral density distribution occurs in the midwave infrared (MIR) wavelength region at 3–5 m.
222
I. Walter et al. / Acta Astronautica 56 (2005) 221 – 229
Table 1 BIRD multi-sensor system parameters (altitude 572 km)
Wavelength Focal length Field of View F-number Detector Detector cooling Pixel size Number of pixels Quantization Ground pixel sizea Swath widtha
WAOSS-B
MIR
TIR
600–670 nm 840–900 nm 21.65 mm 50◦ 2.8 3 CCD-line arrays Passive, 20 ◦ C 7 × 7 m 2 2884 11 bit 185 m 533 km
3.4–4.2 m
8.5–9.3 m
46.39 mm 19◦ 2.0 MCT-line array Stirling, 105 K 30 × 30 m2 2 × 512 staggered 14 bit 370 m 190 km
46.39 mm 19◦ 2.0 MCT-line array Stirling, 80 K 30 × 30 m2 2 × 512 staggered 14 bit 370 m 190 km
WAOSS-B—Wide angle optoelectronic stereo scanner, modified for BIRD, MIR—Medium infrared sensor, TIR—Thermal infrared sensor. a Orbit height = 572 km.
Hence it is the main target for a space-born fire sensor. Reflection effects (e.g. sun glint) as sources of data misinterpretation have to be distinguished by combination of MIR and VIS (visible) channels. Furthermore, temperature information has to be derived from thermal infrared (TIR) imaging data by bi-spectral data processing techniques. The so arising multi-channel system is able to detect hot spots of sub-pixel size.
2.1. Payload architecture and parameters The payload is designed to fulfil the scientific requirements under the conditions of a microsatellite. It consists of the following main parts (Table 1): • Bi-spectral infrared HSRS (HSRS including MIR and TIR-Sensor), • Wide-angle opto-electronic stereo scanner (WAOSSB, VNIR-Sensor) for vegetation analysis and fire classification, • Payload data handling system (PDH) with a mass memory, • A neural network classificator for an on-board classification experiment.
2.2. HSRS—the main BIRD payload The main payload of the BIRD micro-satellite is the newly developed hot spot recognition system (HSRS). Its a dual-channel instrument for middle and thermal infrared imagery based on cooled MCT-line detectors. These linear arrays consisting of 2 × 512 pixels in staggered configuration are based on loophole diode technology with detector electronics on a hybridised detector assembly. Their operating temperatures are 80 K for the TIR and 105 K for the MIR channel, respectively. Based on experiences with a airborne laboratory model using a split Stirling cooler design the detector assembly was optimised regarding stray-light suppression and power needs to meet the radiometric requirements and to cope with the limited resources of the micro-satellite concept. A detector assembly for military applications (see Fig. 1) was modified into a space-qualified solution. The main aspects of the sensor design are: • integrated miniaturised Stirling cycle cooling device, • adapted cold shield design inside the Dewar covered by the IR-window,
I. Walter et al. / Acta Astronautica 56 (2005) 221 – 229
223
Stirling Cooler Motor incl. Drive electronics
Cooled Finger Dewar Dewar Flange incl. Electr. Connectors
Cold Shield Assembly Crank Mechanism Body IRwindow
Focal Plate Assembly incl. MCT Detector
MCT array 2x512
Fig. 1. Miniaturised integrated detector/cooler assembly.
• •
30 x 30 µm pitch, loop-hole diodes
Fig. 2. Design overview of the HSRS detector assemblies.
• special cold finger interfaces for stiffness against dynamical loads with sufficient thermal decoupling. This system consumes 11 W, which is a factor of 4 less in comparison to a split cooler concept. Fig. 2 provides the main constructive aspects of the integrated detector cooler assembly. The design of the flight configuration of the HSRS is driven by strong requirements of the coalignment stability in the line-of-sight of the MIR to the TIR channel (±0.2 arcmin at duty cycle). It is characterised by the following attributes: • identical construction of the optical channels (only optical coating of lenses different, Fig. 5), • autonomous thermal control including heat storage and radiator, • opto-mechanical CFRM-structure with high resonance (> 200 Hz). Fig. 3 provides the overview of the MIR/TIR sensor head design and Fig. 4 gives an impression of the flight hardware completed by thermal control components like radiators and multi-layer insulation. A serious problem of the push-broom sensor, especially in the TIR region is the radiometric stability of the detector elements during the data records which
cannot be covered by laboratory ground calibration. Hence is it essential to add an in-flight calibration device. Thus the HSRS has black body units as temperature reference targets in front of the IR-lenses which can be exposed to the complete field of view at the beginning and the end of the data records. In the passive periods this device serves as protection cover of the optics against contamination and undesirable cool down. The inherent total failure risk is minimised by a separation mechanism that rejects the cover system in the case of emergency. 2.3. WAOSS-B—an example of hardware re-use The WAOSS-B is the VIS/NIR-sensor of the BIRD payload. Actually developed for the MARS-96 mission a flight spare model was modified slightly for the application on BIRD (Fig. 6). A special lens design with integrated filter provides narrow spectral bands over a wide field of view of 50◦ (see Fig. 6). It adopts the instrument to the tasks of vegetation exploration as presented in Section 4. The in-track stereo capability is an important feature for the cloud investigations.
224
I. Walter et al. / Acta Astronautica 56 (2005) 221 – 229
Stirling Cooler Engine Optics Block (Invar) Focusing / Alignment Device
IR-Baffle incl. Cover IF
Optomechanical Structure incl. Heat Storage Detector Proximity Electronics
Cover Separation Device
IR-Dual Band Lenses Inflight Calibration Unit incl. Black Body
Cover Drive Mechanism
Fig. 3. HSRS sensor head design overview.
3. Multi-sensor assembly on a microsatellite 3.1. Structural concept of BIRD
Fig. 4. HSRS with IR electronics unit in flight configuration.
Due to the modular electronic concept of WAOSS-B almost no hardware modification was needed and the reconfiguration is concentrated on the software side, i.e. to provide the master clock function for simultaneous imaging of all BIRD instruments (Fig. 7).
Classical spacecraft designs are oriented on a clear separation of payload which are the scientific instruments and the spacecraft bus containing all subsystem supporting the function of the payload. Here a strong interaction during the design process requires strict management of interfaces from a very early design phase on. In configurations of small satellites this in addition takes place within a small volume. Therefore for BIRD payload and subsystem components had to be developed in a highly iterative process considering the satellite as one complex device. Interface modifications were controlled by defining a modular system for the structure. Thus the BIRD architecture is a cube-shaped tower of three segments (Fig. 8): • service segment as the basic segment with separable launcher interface containing batteries, reaction wheels and gyroscope as well as the GPS system,
I. Walter et al. / Acta Astronautica 56 (2005) 221 – 229
225
Fig. 5. Preintegrated optical bench with the dual lenses and CFRM-structure for thermo-optical stability.
Payload Segment
S-Band Antenna
Electronics Segment Service Segment
Fig. 8. Flight configuration of BIRD (0.6 × 0.6 × 1.6 m3 ).
Fig. 6. Spectral bands of WAOSS-B (NIR doubled).
• electronics segment as a container of payload, spacecraft electronics and the communications package, • payload segment as a platform for the scientific instruments and the star sensors. The configuration is completed by a solar cell system of three panels—two of them deployable—the antennas and magnetic torquers. 3.2. Payload platform concept and technology
Fig. 7. WAOSS-B under integration (8.4 kg, 0.18×0.21×0.38 m3 ).
Due to the power needs of the payload of 200 W in full operation the BIRD mission concept is based on intermediate operation of the scientific instrumentation of approximately 30 min per orbit (10 min data take) and the sun pointing for most effective energy reconstitution. Here the challenge is to cope with the cyclic thermal loads and their influences on the geometric-optical stability of the instruments
226
I. Walter et al. / Acta Astronautica 56 (2005) 221 – 229
Fig. 9. Design principle of the payload platform (cross section).
Fig. 11. BIRD ready for launch as a piggy-back payload on the Indian PSLV C3 vehicle.
Fig. 10. CFRM platform incl. payload mounting points and thermal control elements.
nections with thermal relevance are glued with silverfilled epoxy. The heat conduction reaches over 50% of that from aluminium. Its CTE instead is a factor of 15 less. Another effect is the decoupling of structural loads coming from the spacecraft bus because of separated mounting planes for spacecraft and scientific instruments.
4. In-orbit results line-of-sights which have to be stable within a subpixel range. Early considerations showed the necessity of a thermal control of 1 K for the instrument structure if a conventional isostatic aluminium plate would be used. Despite the large mass of such a solution that was not feasible within the microsatellites resources. The resulting technical solution of the instrument platform is to combine two CF-honey-comb panels on top and below of a 3-mm-thick layer of carbon fibre carbon (CFC). This layer provides heat conduction to the heat pipe interfaces on the front ends of the platform with a coefficient of 155 W m−1 K −1 —comparable with aluminium. The instrument mounting interfaces of this multisandwich structure are inserts with an enlarged plate founded in the CFC-layer (see Figs. 9 and 10). All con-
With the launch on 22 October 2001 BIRD (Fig. 11) was set into a circular sun-synchronous orbit with an altitude of 572 km. Ever since several scientific results regarding hot spot hot detection have been achieved like: • wild fire detection in Brazil (simultaneously with other space-born sensors), • bush fire monitoring in Sydney and Canberra, • investigation of volcanic activities in Chile, Mexico and Hawai, • determination of industrial hot spots in Europe, • coal seem fire investigation in China. The calibration status, especially of the HSRS is monitored by comparisons of temperature data over water
I. Walter et al. / Acta Astronautica 56 (2005) 221 – 229
227
Fig. 12. Vegetation re-growth after fire in New South Wales (Australia, January 2003 [1]): (1) geographical map; (2) MIR image with fire fronts; (3) NIR image with burned areas; (4) NIR image 2 months later.
and during each duty cycle via channelwise imaging of the black bodies. Also a ground-truth measurement campaign in Bavaria in January 2003 was successful. A 2 × 2 m2 sized bonfire (900 K) was lit and measured in parallel
to the BIRD imaging proving the high sensibility of the system beyond its nominal lifetime of 1 year in orbit. Apart from detecting hotspots the images of WAOSS and HSRS can be used to determine the area
228
I. Walter et al. / Acta Astronautica 56 (2005) 221 – 229
affected by the fire allowing e.g. an estimation of the extent and consequences of vegetation fires. If the area is surveyed in regular intervals the vegetation re-growth can be monitored. An example is given in Fig. 12. 4.1. Pixel coregistration aspects The coregistration of the pixels is essential to acquire the accurate values for temperature and area of the hot spots as well as their energy release. Even for hot spots in the sub-pixel range (typ. size down to 12 m2 , T = 1000 K) this is important. Based on the data ground processing which is performed to generate the hot spot characteristics it is possible to derive status information about the payload configuration in space. Table 2 lists typical values, derived from [2] of the pixel displacement of the row data which have to be corrected during the matching process. Here Y is the flight direction while X is the across-track direction. Although tilt effects and time-related influences especially in flight direction cannot be assessed it can be seen clearly that the NIR cannel shows remarkable more dynamics than the TIR channel. Projected to the payload segment design that means that the WAOSS-B as a stand- alone instrument is influenced more than the IR-channels which are
designed as twins on a common opto-mechanical structure. Here the order of magnitude is the same as found in the vacuum in situ tests. A new quality of the BIRD data is its geo-referenced format by means of the correction of all flight attitude data. Here also a measure is given for the coalignment status. However in this case the complete error chain of deviations is covered. This work is under progress but 200 m misregistration typical for the WAOSS-B instrument is reported in [1].
5. Growth potential Although the BIRD payload was developed in a straightforward manner typical for micro-satellites since the origin of its IR-detector technology almost a decade has past. So it is worth to review the latest developments to prove their potential for new applications. Here the recently upcoming combination of two spectral bands sensitivities on a common pixel in the infrared range is of interest. This dualband detectors take advantage of the advances in micro-structures fabrication (see Fig. 13) providing photosensitive areas in the MIR and in the TIR
Table 2 Record of the misalignment correction for ground processing of BIRD data over 8 months Image date
011123 011218 020208 020215 020320 020323 020326 020327 020409 020471 020418 025017 020718 020719
Image target
Europe US China China China China Benin Titikaka Reunion Australia Australia Borneo California Etna
Channel displacement related to the MIR image in pixels NIR (Y, X)
TIR (Y, X)
−4, −1, −5, −4, −5, −5, −7, −2, 0, 0, −4, −6, −4, −5,
0, 0, −1, −1, −1, −1, −1, −1, −1, −1, −1, −1, −1, −1,
2 7 5 4 −9 4 1 6 5 5 6 −2 −10 6
0 0 0 0 0 0 0 −1 −1 −2 −2 −1 −1 −1
Fig. 13. Micro-structured epitaxial layers for 2-colour MCT detector (from [3]).
I. Walter et al. / Acta Astronautica 56 (2005) 221 – 229
229
Fig. 14. BIRD payload platform (460 × 460 mm2 ) with the potential of higher integration by application of modern technology.
Fig. 15. Miniaturised interferometer principle [4]: (1) retro-reflectors, (2) pendulum system, (3) beamsplitter, (4) detector system.
within a pixel pitch. This was demonstrated for quantum-well detectors already and now extended to MCT detectors too. Besides the integration of the spectral bands the sandwich-type configuration provides an excellent coregistration of the pixels. It is by far the best approach to meet the related requirements for the hot spot detection. Projected to the BIRD payload platform this gives the opportunity to integrate the channels of the HSRS into one common opto-electronical system while the exempted part is filled by a modern VIS/NIR-imaging
system. Hence the HSRS would have integrated the functions of WAOSS with the result of having onethird of the payload platform space empty for new instrumentation. This is illustrated in Fig. 14. One idea for that is to add a spot sounding spectrometer for the assessment of gases and aerosols for an enhanced hot spot detection technique. A flight proven design according [4] is a double-pendulum interferometer which provides a robust design with maximised compactness reaching a high optical throughput (see Fig. 15). The described approach would keep the hot spot detection system suitable for small satellites with substantial improvements of its performance. It is at least a contribution to a cost effective and consequently realistic space-borne disaster monitoring system. References [1] E. Lorenz, et al., Scientific and operational exploitation of BIRD, FIREBIRD Technical Notes, Report to ESA, 2003, p. 22 ff. [2] B. Zhukov, et al., BIRD detection and analysis of hightemperature events: first results, Proceedings of SPIE 4886, 2002. [3] W. Ziegler, et al., 3. Generation developments in FPAtechnologies, Company Presentation AIM GmbH, 2003, p. 3. [4] H. Hirsch, Optical design and performance of the planetary fourier spectrometer (PFS), Mikrochim Acta (Suppl.) 14 (1997) 571–574.