Acquisition programming integration of image satellites in LAPAN

Acquisition of satellites images programming become an important issue in LAPAN, due to inaccuracy in acquiring image location. Clouds and smokes cause the low quality of the image and also different time acquisition could make a different synoptic view between different sensor. Some integration of three kinds of sensor has to be created for achieving good location with a high quality image. The other reason why the integration needed that because there is still no research about it. This research aim is to get the best acquisition plan by using three different sensors mechanism like SPOT 6/7, Pleiades and Terra SAR-X. This integration also helps Indonesia government in achieving Government National Program.


Introduction
SPOT-4 programming was the predecessor before SPOT-6 and 7 was launched, SPOT-4 programming was simpler than now. The parameter was used are only base to height ratio, cloud cover (quality of the image), the chance of success, coverage method, survey method, incidence angles, date and area of interest. According to some papers that higher the altitude affected the wider of its swath, if the altitude of satellite is not too high than the coverage is not wide as the higher altitude.
The Table 1 explain about SPOT satellite constellation from SPOT-1, -2, -3, -4 and -5. The other explanation is about bands, spectral range and spatial resolution, those parametric evolve every year. The latest technology is acquired by SPOT-5 because SPOT 6 and 7 did not compare in this table. Panchromatic, red, green, blue, Near Infra Red (NIR) and Short Wave Infra Red (SWIR) are available for those constellation. The highest resolution is come from SPOT -5 which have been interpolated from two sensor into 2.5 meter, and the lowest spatial resolution is 1.15 kilometer (SPOT-4 Vegetation). SPOT-1, -2, -3 and -4 are designed without the blue band, but SPOT-3 failed to launch. The color combination only using red, green and infra red, this cause those image only have false Table 1. SPOT Constellation SPOT-1, -2, -3, -4 and -5 The next table, Table 2 describe the satellite specification for SPOT-6 and 7. The scene store into 12 bits data quantization. NAOMI (New AstroSat Optical Modular Instrument) instrument is a pushbroom imager type which is used by SPOT-6 and -7 Satellite. The spatial resolution of the panchromatic band is range between 1.5 and 2.5 meters and 6 to 10 m for its multispectral band, this condition qualified if the nadir acquisition was achieved. SPOT-6 launched on September 9th, 2012 and SPOT-7 launched on June 30th, 2014, these satellites deployed into the same orbital plane with 180-degree phase. SPOT constellation flies in sun-synchronous orbit type and acquires the image with swath 60 km in width. The altitude of the satellites is 822 km (for SPOT-1,-2,-3,-4 and -5) and 694 km (for SPOT-6 and -7). Mission operator is CNES (for SPOT-1,-2,-3,-4 and -5) and Airbus Defence and Space (for SPOT-6 and -7).

Figure 1. SPOT and Pleiades Constellations
Pleiades 1A is launched on December 17th, 2011 and Pleiades 1B is launched on December 2nd, 2012. Figure 1 describe how those twin satellites flying together in the same orbit with phase 180 degree and along with SPOT constellation. These conditions applies to SPOT and Pleaides satellites, even though each satellite has different time launch they automatically follow programmable orbit. Ground sampling distance (nadir view) is 70 cm on the panchromatic band and 2.8 meters on multispectral but for product resolution its interpolated into 50 cm for panchromatic and 2 meters on multispectral. The swath width is 20 km with an altitude of satellite 694 kilometers. Next figure will visualize how TerraSAR-X and TanDEM-X fly together with altitude 514 kilometers.   Figure 2 display the satellite orbit from two satellites (TerraSAR-X and TanDEM-X orbit). The red and green lines represent orbit for both satellite, the line simulates an elliptical orbit. The color only view the different orbit from each satellite, both lines can be twisted for example red represent TerraSAR-X or vice versa. TerraSAR-X is launched in June 2007 and TanDEM-X is launched in June 2010. They both fly together side by side with helix formation, the distance between them was very close for the flying object it is around 200 -300 meter. Risk of collision is very high for helix orbit with an altitude of 514 kilometers. Single, dual and quadruple polarisation is available to depend on its imaging mode and certain condition. With revisit time along 11 days for acquisition in the same place and orbit inclination in 97.44 degree, those twin SAR satellites prefer to look in the right direction along the track. Next table will be discuss about satellites launching timeline, this timeline has an aim to show the user which one active satellite. It can be seen also the shortest (SPOT-3) and longest timeline (SPOT-2).

Table 3. Satellites Timeline
This table 3 shows that since 2014 from now there are 6 active satellites which can be ordered through programming request. In the future, those six satellites will be integrated into one system or platform that can be managed to achieve optimization result.

Literature Review
The literature started with topic acquisition programming integration between Optic and Radar Constellation. Literature study for the mechanism, procedure, and programming on image satellites, in this case like Optical and Radar imagery. Literature research also covers the satellite orbit definition and how to integrate several image satellites programming in one system. For optic imagery, they are several types of the satellite but it is limited into a high-resolution image. Those high spatial resolution image like SPOT-6 and -7 also Pleiades 1A and 1B. Radar constellation only limited to TerraSAR-X only. Searching also focused on incidence angle and cloud cover term because these terms become major issues for differentiating optic and radar constellation. Radar did not require cloud cover parameter where the optic sensor is dependant to clouds and solar lights. Incidence angle applies for both sensor but the value is set different, the radar of SAR sensor need side looking imaging but optic need vertical imaging (called nadir). Processor gain attenuator is very crucial to the radar constellation but optic does not have an effect too much. Study on predecessor satellite-like SPOT-1,-2,-3,-4 and -5 were regarded important. High spatial resolution change since SPOT-5 was 10 meter become 2.5 meters were a very revolutionary choice. Now the SPOT-6, -7 and Pleiades were having a very high spatial resolution, the blue band also added since SPOT -6 was assembled. Mostly user guide on one access portal, TerraSAR-X, SPOT, and the Pleiades was cited in this research.

Aims
This research purpose is to ensure the availability of optimal and efficient studies of integration from high-resolution imagery, very high-resolution imagery and Synthetic Aperture Radar imagery programming. Also to improve the optimization and operational effectiveness of multi-mission programming stations (the use of the credit, programming level and request time).

Methodology
Literature study for the mechanism, procedure, and programming on image satellites, in this case like Optical and Radar imagery. Optical data like SPOT-6, SPOT-7, Pleiades 1A, and Pleiades 1B. Literature research also covers the satellite orbit definition and how to integrate several image satellites programming in one system.

Concepts and Definition
Most of the scientific satellites usually located in a low earth orbit, six of these satellites also in low earth orbit. These three types of Earth orbit: high, medium and low earth orbit. High earth orbit flies in altitude of 35,780 km above the earth while medium earth orbit has altitude range from 2,000 to 35,780 km. Low earth orbit has a range of altitude between 180 and 2000 km. Those four satellites travel inside Sun-Synchronous orbit, this orbit has a constant angle value for orbit space, thus provide consistent brightness to the satellite. This orbit gives benefit to the earth observation satellite because it is always receiving constant sun illumination. The other orbit like Polar is inclined approximately 90 degrees to equatorial space, this inclination can cover north and south pole area. Incidence Angle is the angle between ground normal and looks direction from satellite, in the next figure the incidence angle represented as β. SPOT 6/7 have ranged in incidence angle from 0-55 degree, and for Pleiades, the range is from 0-33.6 degree. Viewing Angle (which is represented as α) is the angle between look directly from the satellite and nadir. Look direction angle from the satellite may be projected onto two planes defined in the local orbital frame: (yaw axis, pitch axis) and (yaw axis, roll axis).
where IFOV for panchromatic (rad) =1.00E-06 and multispectral (rad) = 4.00E-06. The earth regarded as ellipsoid will have two axes which are semi-major axis (6378.14 km) and semi-minor axis (6356.75 km). The radius of earth represented by R_E, the value of radius is approximately 6367.45 km in average. Table 4 show the result from GSD calculation for several numbers input of global viewing angle (  ).

Table 4. Calculation of GSD PAN and MS
It can be concluded that the higher global viewing angle then the GSD error also have higher number.

Programming Acquisition Comparison
In this part it will be filled some mechanism, procedure and programming of several types of satellites like SPOT-4 in several years behind and comparing to SPOT 6 or 7 also for radar programming (TerraSAR-X).

SPOT-2, and -4
SPOT-4 example was taken on this programming request, Sirius portal and software were used to order SPOT-4 and SPOt-2 image. Proposed images were chosen based on several circumstance and consideration, for example, if the image has high cloud cover but clear for some area, thin haze, perfect incidence angle and another thing. Next figure is the graphical user interface of the Sirius portal.   Programming Pleiades 1A and 1B similar to SPOT-6 and -7 and this programming only described on the Pleiades. Table 5 show the list of proposed acquiring strips for Kukar region (Pemkab Kukar) where incidence angle and cloud cover are shown. Combined, roll and pitch angle also showed here, the quick look can be directly referred from the link to help visualize the image display. rejected, refused and out of specification. Those status created according to certain criteria. Inside one access portal, the Pleiades can be ordered with certain parameters, those parameters like programming level (standard, priority and urgent), acquisition method (mono, stereo, and tri-stereo), Incidence angle (represented positive and negative real number) and validity date. For operational programming, the default of base to height ratio is set by system default 0 to 1 but according to the Pleiades user guide, the number will be various. For example, base to height ratio standard values are 0.4 to 0.7 for stereo and 0.2-0.35 in each pair of tri-stereo image. Optimum base over height ratio really depends on image relief. The explanation detail will continue into the next chapter (Chapter 3. One Access Portal).

TerraSAR-X
The programming can query an order and image database based on certain criteria like sensor mode, polarisation, acquisition range, looking and path direction, range of incidence angle. TerraSAR-X products can be divided into two product groups, those groups are basic and enhanced image product. The basic product image or Level 1-B product group have consisted of EEC (Enhance Ellipsoid Corrected), Single Look Slant Range Complex (SSC), Multi-look Ground Range Detection (MGD) and Geo-coded Ellipsoid Corrected (GEC). While the enhanced image product group are orthorectified, radiometrically corrected, mosaic and ascending or descending mode. The standard cartographic projection for TerraSAR-X product is Universal Transversal Mercator (UTM) and Uniform Polar Stereographic (UPS) with WGS 84 Ellipsoid.  Figure 6. GUI of TerraSAR-X criteria for query Figure 6 show some input parameter for radar programming request, it can be seen that different parameter determined compared to SPOT or Pleiades. Processor gain attenuation parameter has an impact on the scaling process, for example, if the parameter set to 0 dB then it is suitable for the military purpose where shadows of low values are of interest. If the image was analyzed for an urban mapping detection then it will require many high returns where the value of 10 dB. TerraSAR-X product is available in 16-bit integer values, sometimes very bright target (like corner object) exceed 16-bit data range. A value scaling is required to delete those high value on a very bright object, the processor gain attenuation will increase or decrease the radiometric contrast. Terra SAR product order can be set under user requirement like processing level (for example L1B), product type (EEC, GEC, SSC, MGD), orbit precision, processing priority, projection, customer acquisition priority, processor gain attenuation and other.

One Access Portal
One access portal (OAP) is a system that integrates all airbus product like SPOT-6/7, Pleiades-1A/1B, and TerraSAR-X (and TanDEM-X). This system can be used for ordering images by handling certain parameters, it also can visualize orbit pass of each satellite. Report on successful image acquisition can be obtained through the portal on the web format.

Interface
The graphic user interface from one access portal can be view by accessing the web page address https://www.intelligence-airbusds.com/, here is the figure look-alike. After login with the username and password, the web page will automatically refer to the tasking cockpit on Airbus desktop.

ICR and Programming Request
Imagery Customer Request (ICR) is the name of order identification, request programming for Pemalang Regency Center of Java Province (ICR number 227412) can be seen in the next figure. The next figure is displayed several scenes of Pleiades quick look where it is overlaid with the area of interest, there also exist a list of several acquiring strips.

Optimization Parameters
Each parameter from programming acquisition input has an important role in making the high quality of image, the result of optimization has to be combined from several parameters or constraints. For example like the base to height ratio, cloud cover, incidence angle, area of interest and others.

Stereo (Base to Height Ratio)
In aerial photography base height ratio is defined as "the distance on the ground between the centers of overlapping photos, divided by aircraft altitude".  figure 9, it can be concluded that a higher ratio of the base over height will not cover an object in the valley which is located between two mountains. The baseline is actually created by a distance between two locations of the different satellite (or same satellite but in the different acquisition of timing). The height that defined here is the distance between perpendicular baseline into earth surface or the altitude of the satellite.

Cloud Cover
Cloud cover is the coverage of valuable clouds present in satellite imagery which are usually described as indexes, percentages, scores, etc. For SPOT-4 and SPOT-2 it is usually represented by a score like (AAAAB) but the values are different for SPOT-6 and SPOT-7. The percentage represented from 0-100 percent if 0 percent cloud cover then the display will be a clear picture. Thin fog, gloomy pictures (cirrus or Columbus nimbus), smoke (caused by fire) and acquisition of errors can be one of the factors that cause the quality of cloud cover. This parameter is not required for TerraSAR-X programming requests.

Area of Interest
Area of Interest is an area created by the user to select imageries which intersect with the user creation shape. The shape coordinate can be determined by the user through the web (on the fly) and uploading from an internal computer. According to SPOT 6 user guide that AOI is "An AOI outlines a particular region by panel, shape, preset values, or by a defined line and sample". Area of interest is based on a user request for example mountain area, hilly, lake, sea, the forest is a different object which determines the request. Disaster usually locates the AOI randomly, but for the monitoring like forest, mangrove, peat-land, paddy field monitoring are repeating the request with the same AOI.

Incidence Angle
Incident angle for SPOT-6 and -7 range between 0-55 degree and for the Pleiades ranging from 0 until 33.6 degrees. But for ordering Pleiades or SPOT the range will be limited from 0 -30 degree. Incident angle for TerraSAR-X ranging from 15 until 60 degrees but for best performance of the quality image the range will be display on the next table. The incident angle on Radar Constellation will various and it depend on imaging mode and polarization mode. Table 6. Recommended Performance Beams and Incidence Angle Ranges for High Contrast Scenes Table 6 is explain about range value of incidence angle or look angle for every radar mode of acquistion and also its polarization mode. Zero value on incidence angle is the perfect or best acquisition from the optic sensor, the incidence angle approach zero will benefit to nadir view imaging. But this zero's incidence is not also applied to the radar sensor, because all energy that transmits from the radar sensor to the object in the ground will perfectly reflect back to the sensor. This condition makes reflection was accumulated and the intensity becomes high, the resulting image will bright in all area.

Date Acquisition
The time range is an important parameter to support the chance rate success result. If the range is too short then it will produce failure result and if a long period was chosen then it will provide an error result. This basic and primary parameter is an important key to the programming request, mission plan of image acquisition.

Credit Accounting
The provider gives some concept in calculating the requested order (acquisition amount/unit) which are different for every sensor. In this part, there will be some understanding of how a user can calculate how much order or amount that can be acquired. One credit is equal to one "ES" or called Equivalent Scene, limitation of one-year consumption will be maximum to 3000 credits. The contract between SPOT and Pleiades will be differentiated, for SPOT there will no credit account. Unlimited consumption on SPOT Constellation or open access but only limited to priority and urgent request programming for 100 in number annually. Pleiades programming request will be limited to 3000 credit where 1 credit equal to 1 equivalent scene (called "ES").

Pleiades
Pleiades image has 400 km square area from 20 km (relative upon its incidence angle) for each side. For 20 km in length Pleiades sensor require 3.125 seconds to accomplish its acquisition

SPOT
SPOT-6/7 sensor has a length and width 60 km for each side, and it means nearly 3,600 km square for its area which is equal to one credit. From the last figure, it is shown that Pleiades need 0.99 ES to establish one mesh but for the SPOT it only requires 0.94 ES. The SPOT images are not limited for acquiring the data, the ES credit calculation did not affect much for SPOT programming request this year.

TerraSAR-X
Credit calculation on TerraSAR-X is different from SPOT and Pleiades Constellation because there are several imaging modes. Wide ScanSAR, ScanSAR, Strip Map, SpotLight and High-Resolution Spot Light, there are can be 1500 square kilometer equivalent. For example Strip Map (3m) square kilometer scene size is 30 km in width multiply with 50 km in length which is equal to 1500 kilometer square. But if the example for Wide ScanSAR (40m) the length 270 km multiply with 200 km then the result is 54000-kilometer square which is not equal to 1500 km2. The value 54,000 km2 will be equivalent to 1,500 km2 from Strip Map, the equivalence will be displayed in the next table.  Table 7. TerraSAR-X Credit Accounting Table 7 can conclude that square kilometer which acquired by scene size will change for every 600 credit value. For example, if all credit utilized only for wide ScanSAR then 600 credit units multiply with 270x200 will equal to 32,400,000 kilometer square. Compare to all 600 credit apply only to HS SpotLight then the calculation will be 600*10*5 or equal to a 30,000-kilometer square.

Conclusion
Optic and radar images are different beside from its sensor, but also for choosing incidence angle, cloud cover. For a radar image, one does not need to set how many clouds cover percentage because radar sensor penetrates clouds and small particle of rains. If incidence angle value in the optic sensor approach to zero then it will have a good image but this condition is opposite to radar. Radar sensor should not equal to zero or even approach to zero because the object will reflect back perfectly, this condition causes high intensity on the image and results to the bright image on all area. That is why the value of radar incidence angle started from non-zero value for example 15 or 17 degrees.
Chance of the programming request result is really determined by that parameter, and if an operator chooses the wrong parameter then the result will be low or completely failed. For example, revisit time parameter can affect date acquisition on a specific location, if TerraSAR-X can acquire an image of Jakarta on January 1st, 2025 and the period of acquiring image set between January 2nd, 2025 until January 9th, 2025 then this will not produce any TerraSAR-X image because the revisit time of TerraSAR-X is 11 days long. On another example like if someone put cloud cover on 0 percentage or zero on incidence angle value of an optic image then the chance of the result will be too low or fail.
From this research exploration that one can see the different procedure and parameter from radar and optic satellite. Those difference hopefully can be integrated into an efficient integrated acquisition programming method. By using consideration of radar and optic parameter acquisition which have been exposed in this research. For example the different of an incidence angle value between radar and optic, it can determine the optimal incidence angle for each sensor.