SAR satellite configuration for determining the geo-coordinates of radar point targets on the earth's surface and method using such a satellite configuration

Abstract

The satellite configuration has at least two first SAR satellites for interferometry and at least one second SAR satellite for radargrammetry. In each case two of the first SAR satellites have an across-track distance of 10 m to 1 km. The at least one second SAR satellite forms a radargrammetrically effective across-track distance of 100 km to 1000 km with each of the first SAR satellites. The measured values of the interferometric measurement can be calibrated by means of the measured values of the radargrammetric measurement. All first and second SAR satellites have the same illumination direction with respect to the orbit direction. The illumination direction of all first and second satellites is either left-looking or right-looking at the same time.

Description

[0001] SAR satellite configuration for determining the geo-coordinates of radar point targets on the Earth's surface and methods using such a satellite configuration

[0002] The invention relates to a SAR satellite configuration for remote sensing of the Earth's surface, in particular for determining the geo-coordinates of radar point targets, as well as a method for remote sensing of the Earth's surface using SAR satellites with the aforementioned SAR satellite configuration and a method for processing the data from the images of the SAR satellites of such a SAR satellite configuration for forming the geo-coordinates of radar point targets on the Earth's surface.

[0003] The object of the invention is the precise determination of the geo-coordinates of prominent and / or specially designed (in this respect "artificial") radar point targets on the Earth's surface with high density and with an accuracy in the decimeter (dm) range using a configuration of SAR (Synthetic Aperture Radar) satellites. The object of the invention is also to be able to provide the coordinates of objects on the ground within a short time after the satellites have passed overhead.

[0004] The satellite systems used so far require a longer period of time to acquire the necessary data, as several satellite passes are required, which may be several days apart.

[0005] The invention includes a method for establishing satellite orbits for a total of three or more Earth observation satellites, of which at least two satellites are in close formation (i.e., typically with distances on the order of 10). 1 m to 10 3The system consists of two satellites in close formation, as well as at least one other satellite at a greater distance from the formation (i.e., several hundred kilometers away), with all participating satellites observing ("recording") a specific area on the ground as synchronously as possible. This combination of satellites in close formation and at least one satellite flying at a greater distance is referred to below as a satellite configuration. To determine the geographic coordinates of objects on the ground in near real-time using a SAR satellite configuration, a close formation flight of two satellites is necessary (for performing SAR interferometry, as in the TanDEM-X project), as well as a third satellite that is located far away (several hundred kilometers from the other two satellites) and, in conjunction with the other satellites, enables radargrammetry.The task is therefore to synchronize a widely distributed satellite configuration so that it can observe an area on the ground as simultaneously as possible, or with a small time offset (seconds to a few minutes).

[0006] With optical satellites equipped with a camera for image capture, this is relatively easy to a certain extent by aligning the cameras with the target area. This is done either by controlling the position of the camera itself or by aligning the entire carrier satellite.

[0007] This alignment of the viewing direction for image acquisition is only possible for SAR satellites within a very limited angular range. To facilitate the processing of SAR data, the viewing angle control is typically used to center the received radar echoes on "Doppler zero" (yaw steering). Therefore, SAR satellites can only observe the ground within a narrow angular range. The initial goal of observing a ground area from widely separated orbits as simultaneously as possible can only be achieved with SAR satellites using specially configured orbits.

[0008] Further prior art is known from ROSSI, Christian [et al.]: TanDEM-X calibrated raw DEM generation. In: ISPRS journal of photogrammetry and remote sensing, Vol. 73, 2012, pp. 12-20. - ISSN 0924-2716 and ZINK, Manfred [et al.]: The TanDEM-X mission: overview and status. In: IEEE international geoscience and remote sensing symposium, 23-28 July 2007, pp. 3944-3947. - ISBN 978-1-4244-1211-2. The object of the invention is the precise determination of geo-coordinates of prominent radar target points on the Earth's surface with high density and with an accuracy in the decimeter (dm) range using a configuration of SAR satellites. The aim of the invention is also to be able to make the coordinates of objects on the ground available within a short time after the satellites have passed overhead.It should be noted that the satellite systems used so far require a longer period of time to acquire the necessary data, as several satellite passes are necessary, which may be several days apart.

[0009] To solve this problem, the invention proposes a satellite configuration for remote sensing of the Earth's surface, in particular for near real-time determination of the geo-coordinates of radar point targets, wherein the satellite configuration is provided with at least two first SAR satellites and at least one second SAR satellite, wherein the at least two first SAR satellites are provided for interferometric surveying of a survey area on the Earth's surface and each pair of the first SAR satellites has an cross-track distance of 10 m to 1 km, wherein the at least one second SAR satellite forms a radargrammetrically effective cross-track distance of several hundred km with each of the first SAR satellites.wherein the at least one second SAR satellite is provided with one of the first SAR satellites for radargrammetric surveying of the survey area surveyed by the at least two first SAR satellites, wherein the measured values ​​of the interferometric survey can be calibrated using the measured values ​​of the radargrammetric survey, wherein all first and second SAR satellites have the same illumination direction with respect to the orbital direction, and wherein the illumination direction of all first and second satellites is simultaneously either left-looking or right-looking. In an advantageous embodiment of the invention, it can be provided that the at least two first SAR satellites move at the same average speed in different, nearly circular orbits at the same average altitude.where the orbits differ in the orbital elements eccentricity and / or argument of perigee, creating a vertical cross-track distance between the satellites that varies periodically in length over one orbital revolution, so that areas of interest on Earth (e.g., 20–55° latitude) can be interferometrically measured with a sufficiently effective baseline.

[0010] Furthermore, according to the invention, it may be advantageous for the at least two first SAR satellites to move at the same mean speed in different, nearly circular orbits at the same mean altitude, wherein the orbits lie in different planes, i.e., have different lengths of the ascending node and / or different inclinations, thereby creating a horizontal cross-track distance between the satellites, the length of which varies periodically over one orbital revolution, so that areas of interest on Earth (e.g., 20–55° latitude) can be interferometrically measured with a sufficiently effective baseline.

[0011] As an alternative to the aforementioned embodiments, it can be advantageously provided according to the invention that the at least two first SAR satellites have a vertical and horizontal cross-track offset with the same or different amplitude, so that areas of interest on Earth (e.g. 20 - 55° geographic latitude) can be interferometrically measured with a sufficiently effective baseline.

[0012] In a further advantageous embodiment of the invention, the at least two first SAR satellites may, in addition to a vertical and / or horizontal cross-track offset, have an along-track offset caused by different mean anomalies of the two orbits. Furthermore, it may be advantageous for the at least one second SAR satellite to move with the same mean velocity relative to the at least two first SAR satellites on a different, nearly circular orbit at the same mean altitude, wherein the orbits lie in different planes, i.e., have different ascending node lengths and / or different inclinations, thereby creating a horizontal cross-track distance between the first and second SAR satellites that varies periodically in length over one orbital revolution, so that areas of interest on Earth (e.g.,20 - 55° geographical latitude) with a sufficiently effective radargrammetric Across-Track Baseline.

[0013] Alternatively, it may be advantageous for the at least one second SAR satellite to have, in addition to the horizontal cross-track distance, a vertical cross-track offset relative to the at least two first SAR satellites, caused by differences in the orbital elements eccentricity and / or argument of perigee.

[0014] It is also advantageous if the at least one second SAR satellite precedes or lags behind the at least two first SAR satellites (typically by a few minutes of flight time), which is caused by different mean anomalies, so that the radargrammetrically effective Across-Track Baseline follows from the displacement of the observation target due to the Earth's rotation between the overflight times of the first and second SAR satellites.

[0015] A further embodiment of the invention is characterized by at least one further second SAR satellite which provides an Across-Track Baseline of 10 to each of the other second SAR satellites. 1 m to 10 3 m, each of which has an Across-Track baseline of 10 with each of the first SAR satellites 5 m to 10 6m forms, wherein the first SAR satellites and the second SAR satellites change their functions twice per orbit, in that in the ascending orbit the first SAR satellites are intended for interferometry surveying and at least one of the second SAR satellites or one of the second SAR satellites together with at least one of the first SAR satellites or with at least one other of the second SAR satellites is intended for radargrammetry surveying and in the descending orbit the second SAR satellites are intended for interferometry surveying and at least one of the first SAR satellites or at least one of the first SAR satellites together with at least one of the second SAR satellites or with at least one other of the second SAR satellites is intended for radargrammetry surveying.

[0016] The problem according to the invention is further solved by a method for determining the geo-coordinates of radar point targets, such as radar point spreaders, wherein in the method a satellite configuration according to the invention or one of its previously mentioned embodiments is provided, an interferometry survey of a survey area on the Earth's surface is carried out using the first SAR satellites, a radargrammetry survey of the survey area is carried out using at least one second SAR satellite together with one of the first SAR satellites, for which an interferometry survey has been carried out using the first SAR satellites, and the measured values ​​of the interferometry survey are calibrated using the measured values ​​of the radargrammetry survey.

[0017] In the method according to the invention, it can advantageously be further provided that the satellite configuration is supplemented by at least one further second SAR satellite, which provides an Across-Track-Baseline of 10 to each of the other second SAR satellites. 1 m to 10 3 m, each of which has an Across-Track baseline of 10 with each of the first SAR satellites 5 m to 10 6m forms, and that the first SAR satellites and the second SAR satellites change their functions twice per orbit, whereby in the ascending orbit the first SAR satellites are intended for interferometry surveying and at least one of the second SAR satellites or one of the second SAR satellites together with at least one of the first SAR satellites or with at least one other of the first SAR satellites is intended for radargrammetry surveying and in the descending orbit the second SAR satellites are intended for interferometry surveying and at least one of the first SAR satellites or at least one of the first SAR satellites together with at least one of the second SAR satellites or with at least one other of the second SAR satellites is intended for radargrammetry surveying.

[0018] The method according to the invention enables the determination of highly precise coordinates of isolated radar objects, typically retroreflectors that appear as point-like objects (so-called point scatterers) in the radar image. Such reflective properties are generally only found in "radar reflectors," such as those formed by building corners oriented towards the sensor or by masts arranged perpendicularly on a flat surface. This type of object is typically only found in populated areas. To use the method in areas where such objects are not present, such as deserts, rainforests, tundra, savannas, bodies of water, or alpine regions, corner reflectors, for example, can be strategically deployed.

[0019] With optimally suited radar point spreaders, appropriate data processing can achieve an accuracy of the geo-coordinates in X, Y, and Z in the centimeter range. This value is generally far below the geometric resolution of the SAR system. With less-than-perfect radar point spreaders, such as those found in vehicles, an accuracy in the range of several decimeters can still be achieved.

[0020] The applications of the invention generally lie in determining the geo-coordinates of objects on the Earth's surface and are also applicable whenever a widely distributed satellite configuration is to observe an area on the ground as synchronously as possible. With optical systems, it is often helpful to be able to view an object or scene from different perspectives. In this way, 3D views of the scene, or more generally, so-called surface models, can be generated. The method described here then makes it possible to also create three-dimensional representations of dynamic objects (e.g., vehicles).

[0021] Another application lies in the simultaneous observation of an area or object (e.g., a ship) on the Earth's surface using radar and optical satellites. Synchronizing distant SAR satellites enables bi-static ground-based imaging by utilizing "forward scattering" (radar scattering from the ground towards the passive SAR satellite).

[0022] The method according to the invention makes it possible for operators of SAR and optical satellite formations to carry out these additional applications without additional effort. Up to now, the formation has generally only been used to reduce the so-called "revisit time" (which refers to the shortest possible time interval between two satellite images).

[0023] Precise positioning worldwide has so far only been possible with a GNSS (Global Navigation Satellite System), such as GPS, Galileo, or GLONASS. These systems determine the coordinates of a navigation receiver, which must be deployed on-site. In contrast, the inventive method is based on radar remote sensing techniques that require no technical equipment, i.e., no special infrastructure, in the area of ​​interest.

[0024] The Ground Control Points (GCPs) obtained using the inventive method can be used as reference points for the precise georeferencing of other image data, e.g., aerial or satellite imagery or so-called "mobile mapping data". Many features, such as road markings, can then be extracted from this data, so that the GCPs generally contribute to accurate mapping.

[0025] Furthermore, GCPs can serve as orientation points for vehicles or people in so-called "landmark navigation".

[0026] The water levels of lakes, rivers or floodplains can also be determined using the method according to the invention, provided that masts are placed in the water which, together with the water surface, form a so-called "corner reflector", as described, for example, in WO 2011 / 092056 A2, DE 10 2010 001 440 Al and DE 10 2016 208 508 Al.

[0027] By comparing currently determined 3D positions of objects with previous images, surface shifts, uplifts, or subsidences, e.g., due to oil and gas extraction, as well as changes caused by tectonic events, can be measured. This new method is particularly helpful because it delivers results very quickly and, for example, tectonic shifts can be determined shortly after an earthquake using an image from the satellite configuration, provided older images of the region are available as a reference.

[0028] The system is superior to so-called Persistent Scatterer Interferometry (PSI) in that it directly measures absolute displacements in all three degrees of freedom and not just relative displacements in the "line of sight" of the radar.

[0029] Repeated measurements in urban areas reveal changes in the city's topography, which in turn allows conclusions to be drawn about construction activity.

[0030] Since the coordinates of point targets are recorded very precisely, the system can also be used for energy change detection of such objects. For example, temporarily parked vehicles can be easily distinguished from fixed objects such as lampposts. Using two time-shifted images of the satellite configuration and an analysis of the initially determined object positions, it can be determined whether a change has occurred.

[0031] To solve the aforementioned problems, the invention further provides a method for processing the images from SAR satellites of a satellite configuration comprising a first and a second interferometry SAR satellite and at least one radargrammetry SAR augmentation satellite for forming the geo-coordinates of radar point targets (GCPs) on the Earth's surface, wherein in the method a) interferometry images with the 2D coordinate grids Ai, A2 are provided by the two interferometry SAR satellites and a radargrammetry image with the 2D coordinate grid Bi is provided by the at least one radargrammetry SAR augmentation satellite, b) the image coordinate grids of the first interferometry SAR satellite are geodetic corrected to form first corrected interferometry image coordinate grids AI GECO,a) the image coordinate grid of the second interferometry SAR satellite is processed by geodetic correction to form a second corrected interferometry image coordinate grid A2_GECO, and the image coordinate grid of at least one radar grammetry SAR supplementary satellite is processed by geodetic correction to form a corrected radar grammetry image coordinate grid BI_GECO; c) objects with point-like radar backscatter are detected in the SAR image of the first interferometry SAR satellite, and their coordinates with respect to the first corrected interferometry image coordinate grid AI_GECO are determined, forming a first 2D point set PAI_GECO; d) objects with point-like radar backscatter are detected in the SAR image of the radar grammetry SAR supplementary satellite, and their coordinates with respect to the corrected radar grammetry image coordinate grid BI_GECO are determined, forming a second Form 2D point set PBI_GECO,e) based on the distance between the image coordinates of the points from the first 2D point set PAI_GECO and those of the points from the second 2D point set PBI_GECO, the point pair set MAIBI_GECO is determined, comprising those pairs of points contained in the first and second 2D point sets PAI_GECO and PBI_GECO, respectively, whose mutual distance is below a specified threshold; f) using a stereo-SAR method based on the parallax of their image coordinates, the elements of the point pair set MAIBI_GECO are combined to form a set of 3D coordinates GCPAiBi_stereo of the target points with an estimated accuracy; g) from the set of 3D coordinates GCPAiBi_stereo, those coordinates that have an accuracy better than a specified threshold are grouped into a subset GCPcaiibration of the target points with the most accurate 3D coordinates; h) the two geodetically corrected Interferometry image coordinate grid AI GECO,AZ GECO by estimating their position-dependent shifts relative to each other in Range A, rg and azimuth A azare mapped onto each other, i) then, based on this mapping, the second corrected interferometry image coordinate grid A2_GECO is resampled to form a second corrected interferometry image coordinate grid A2_GEco_Res that is geometrically congruent (co-registered) to the first corrected interferometry image coordinate grid AI_GECO, j) the phase values ​​of the complex-valued interferometry images in the first and in the co-registered second corrected interferometry image coordinate grid AI_GECO, A2_GEco_Res are combined by interferogram formation to form the ambiguous interferometric phase <|) wrapped, k) the interferometric phase <|) unwrapped, freed from ambiguity, is formed by phase unwrapping.l) by phase calibration, both a possible offset of the interferometric phase ^unwrapped to the elevation values ​​of the points GCPcaiibration converted into phase values ​​(and thus to actual elevations) and, by using the GECO phase difference AIGECO formed from the geodetic corrections ARAI_GECO and ARÄ2_GECO converted into phases, image position-dependent, individual geodetic phase errors are compensated, and the calibrated phase <|) unwrapped_calib is obtained; m) the calibrated phase <|) unwrapped_calib is converted into a digital elevation model (DEM) by a method for generating a digital elevation model (DEM); n) by means of a geocoding procedure based on the generated digital elevation model, all of the radargrammetry image points PBI_GECO not used in step g) in the set of geodetically corrected, selected points are directly georeferenced to a set of 3D GCP coordinates. GCPBI_GEOCODED,and o) the quantities GCPBI_GEOCODED and GCPCALIBRATION are combined to form the final product.

[0032] It is known that the coordinates of an object in a SAR image can be mapped onto a 3D geographic coordinate system using a process called "geocoding" with the aid of a digital terrain model (DTM). The problem with this method is that a sufficiently accurate DTM is not available for all areas of the Earth's surface. Due to the oblique viewing angle inherent in the SAR process, the height and position of a pixel are linked, meaning that during geocoding, a height error in the DTM is converted into a positional error of the same order of magnitude. Therefore, to obtain geographic coordinates with an accuracy better than 1 dm, a DTM with a height error better than 1 dm is required. However, if the DTM no longer reflects the current state of the radar-reflecting surface at the time of the radar point measurement, a radar object would be located at the wrong 3D position.The method according to the invention avoids this disadvantage, since all data are recorded in real time and practically simultaneously.

[0033] Radargrammetry methods are also known, in which a ground-based object is measured from two angles (from two overflights from different "orbits") using a large baseline, and its 3D coordinates are determined by means of two precise distance measurements and cross-sectional analysis (see, e.g., [1]). When radargrammetry is used as the sole method, the yield or density of objects that can be measured with high precision on the ground is limited.

[0034] To achieve high accuracy, radargrammetry requires a large baseline (e.g., 100 to 300 km) so that ground-based objects can be observed from significantly different angles. However, the objects being measured typically exhibit high variance in radar backscatter when illuminated with different aspect and incidence angles. Therefore, it often happens that an object is clearly imaged by radar from one view but barely visible from a different, second view. This severely limits the number of objects that can be stereoscopically measured in this way. While a shorter baseline can make the backscatter more similar, the gradual intersection and residual inaccuracies in distance measurement result in large elliptical errors in the 3D localization.This disadvantage is eliminated in the proposed method by subsequently densifying the GCPs using the generated high-precision DGM with the radar detections that were seen from only one viewing angle.

[0035] Methods for determining terrain height using SAR cross-track interferometry are known. In the interferometric method for height measurement, a comparatively small baseline is used (e.g., in the range of 100 to 200 m) compared to the radargrammetric method, thus avoiding the problem of the angle-dependent backscattering properties of an object. However, with interferometric methods, the absolute height of the object cannot be determined without additional information. Furthermore, the determined interferometric phase is ambiguous modulo 2π, so that the heights are initially only known as the modulo of a value derived from the satellite configuration. While this is overcome by so-called "phase unwrapping," absolute height calibration is still necessary.Interferometric processing and DEM generation can, in principle, be used to determine local terrain height and thus for geocoding individual objects. However, as already mentioned, the radar-generated DEMs (e.g., from the TanDEM-X mission) generally do not meet the required height accuracy of "better than 1 dm" for this application. Furthermore, other sufficiently dense reference point networks are not available worldwide with comprehensive coverage and a suitable sampling grid.

[0036] The invention makes it possible in a simple way to simultaneously improve the accuracy of determining the geo-coordinates of radar point targets, increase their yield (density), and enable near real-time acquisition and processing.

[0037] To solve this problem, the invention proposes a satellite configuration for remote sensing of the Earth's surface, in particular for determining the geo-coordinates of radar point targets, such as radar point spreaders, wherein the satellite configuration is provided with three or more satellites.

[0038] The invention is explained in more detail below with reference to the drawing. Specifically, the drawing shows:

[0039] Fig. 1 shows a configuration of three SAR satellites for quasi-simultaneous interferometric and radargrammetric measurements.

[0040] Fig. 2 shows a representation of the Earth-fixed ascending orbits of a satellite configuration for simultaneous interferometric and radargrammetric measurements.

[0041] Fig. 3 shows the acquisition geometry at 49 degrees latitude for the configuration shown in Fig. 2.

[0042] Fig. 4 shows the course of the radargrammetric baseline as a function of latitude for the configuration shown in Fig. 2, Fig. 5 shows the course of the radargrammetric aperture angle as a function of latitude for the configuration shown in Fig. 2,

[0043] Fig. 6 shows the course of the effective baseline of the interferometric formation as a function of latitude for the configuration shown in Fig. 2.

[0044] Fig. 7 shows a representation of the orbits according to Fig. 2, supplemented by an alternative configuration of the radargrammetric satellite with which the ground targets are measured at minute intervals for interferometric recording, and

[0045] Fig. 8 shows a flowchart of the process steps for processing the data to determine geo-coordinates, referring to the special case of two interferometric and one radargrammetric satellites (processing of a radargrammetric pair).

[0046] By combining the two methods of "geocoding using InSAR heights" and "determination of radargrammetric reference heights," the coordinates of significantly more objects on the ground can be determined than with either method separately. This is achieved by the method according to the invention and, in particular, by an optimized satellite configuration of at least three radar satellites.

[0047] Furthermore, the new combined method is significantly more precise because the phase calibration of the cross-track interferometry no longer relies on data from potentially other sensors (such as the LiDAR of the ICESat mission) with different ground penetration depths and scanning grids, but rather uses the SAR data itself. For the calibration of the InSAR phase, objects are used whose position and height can be determined very accurately using radargrammetry. These are typically radar point scatterers, such as those formed by radar reflectors. Advantages of this approach compared to known methods: a) Unambiguous assignment of calibration data during InSAR phase calibration

[0048] The satellite configurations provided according to the invention deliver interferometric and radargrammetric data pairs. The interferometric processing provides the initially ambiguous and uncalibrated height of an object on the ground, while the radargrammetric (stereoscopic) processing provides the absolute object coordinates in 3D space. For the calibration of the interferometric phase, the highest quality point measurements from the radargrammetric processing are selected as calibration data. These are typically the radar scatterers with the best signal-to-clutter ratio (SCR), i.e., the ratio of the signal from the radar backscatter of the point target to the signal from objects in the immediate vicinity.

[0049] To minimize the risk of confusion when combining (i.e., "matching") objects from radiometric and interferometric processing, objects that are geometrically close to each other should be avoided. Since one of the datasets from the three satellites is used in both processing branches, pinpoint accuracy of the objects is ensured during the calibration of the interferometric phase. As already mentioned, this is not always guaranteed when using external reference coordinates, such as the laser height measurements from ICESat. b) High-precision phase calibration information

[0050] For phase calibration using absolute heights from radiometric radargrammetry processing, the point targets with the best signal-to-clutter ratio are selected. These objects, which appear as bright point targets in the radar image, have a much lower phase noise than area targets. Therefore, there is a chance of achieving the required height accuracy of 1 dm not only for the point-like calibration information but also for the area-based height information determined from interferometry. c) Implementation of consistent time-of-flight corrections for both the radargrammetric and interferometric measurements.

[0051] In the inventive method, geodetic correction methods for the travel time of the radar pulses in the ionosphere and atmosphere, as well as for the Earth's tides, are applied to both the radargrammetric processing and the interferometric phases ("Geodetic InSAR Processing"). An advantage of the proposed method is that the near-simultaneous acquisition of all data allows for the application of a temporally and spatially consistent correction model.

[0052] This process corrects the position of the image pixels in range and azimuth. However, since the phase difference of an interferometric pair is also only a (potentially ambiguous) measurement of a distance difference, it can also correct the phase information used in interferometric analysis to determine terrain height.

[0053] The atmospheric error in the interferometric phase during simultaneous acquisition of a satellite formation arises because, in the acquisition with the shallower angle of incidence, the radar signal travels a longer path through the atmosphere and thus experiences a greater signal delay than the radar signal in the acquisition with the steeper angle of incidence. This effect significantly impacts the interferometric phase even with a relatively small cross-track baseline and a correspondingly small difference in the angle of incidence. In contrast to the method according to the invention, previous methods, such as repeat-pass interferometry, require the corrections estimated for each of the two different acquisition times [2], meaning that the atmospheric state description (measurement or model) necessary for the correction is required separately for each time point.In the method according to the invention, this information is only required for a single, short period of time.

[0054] The following description refers to the description of various satellite orbits according to the invention for the interferometry satellites and the at least one radargrammetry satellite. The orbit descriptions are based on the fundamental principles of such orbital descriptions as mentioned, for example, in DE 101 32 723 B4. The content of that patent specification is hereby incorporated by reference into the subject matter of the present patent application.

[0055] To enable the determination of the geo-coordinates of objects on the Earth's surface within a short time after overflight, radar-satellite configurations are proposed according to the invention, with which interferometric and radargrammetric measurements are performed almost simultaneously. The configurations consist of at least three radar satellites, of which at least two are used for the interferometric measurements in a relatively close formation with cross-track distances on the order of 10 1 m to 10 3 flying in m (hereinafter referred to as "interferometry formation"), as well as at least one other radar satellite at a horizontal distance of approximately 10 5 m, which provides a data set for a radargrammetric measurement (hereinafter referred to as the "radargrammetric supplement satellite").

[0056] The following explanations serve as background information and to clarify definitions of various aspects of satellite orbits.

[0057] The spatial distance between two satellites is referred to as the along-track distance along the flight direction and as the across-track distance in the plane perpendicular to the flight direction. The across-track distance consists of a vertical component and a horizontal component.

[0058] In the plane perpendicular to the flight direction, the effective baseline is also defined as the orthogonal distance from one interferometric satellite to the line of sight of the other interferometric satellite. The effective baselines required for interferometric measurements are frequency-dependent.

[0059] X-Band: 100 m ... 1000 m

[0060] C-band: 260 m ... 2600 m

[0061] L-band: 700 m ... 7000 m

[0062] The radargrammetric baseline corresponds to the distance between the interferometric formation and the radargrammetric supplemental satellite.

[0063] It is independent of the frequency band. Order of magnitude: 10 5 m varies with latitude (see Fig. 5). Example for a 500 km orbital altitude: 250 km distance for observations at 55° latitude; 410 km distance at 20° latitude

[0064] The orbit of a satellite around the Earth (or "satellite orbit" for short) is characterized by the following six orbital elements:

[0065] The semi-major axis a and the eccentricity e describe the shape of the orbit. The invention utilizes low, near-circular orbits with a = 6700 ... 7500 km and e < 0.01. The mean orbital altitude is the difference between the semi-major axis and the Earth's radius (6378 km at the equator). The orbit's position in space is described by the inclination i (angle of inclination between the z-axis of the geocentric-equatorial coordinate system and the orbit's angular momentum vector), the length of the ascending node Q (angle between the x-axis of the geocentric-equatorial coordinate system and the line of ascending node), and the argument of perigee co (the angle between the line of ascending node and the perigee radius). Thus, the invention primarily uses sun-synchronous orbits with inclinations i = 97 ... 101°. However, the invention is not limited to this inclination range. The elements Q and co are not restricted (0...360) 0 ).

[0066] The satellite's position in its orbit can be described by the mean anomaly M (the angle between the perigee radius and the satellite's position vector). One revolution corresponds to a change in M ​​of 360°. Therefore, the invention is not limited to a single value for M, but can be used between 0° and 360°. We use the parameter M to describe the separation of two satellites along their orbits: the difference in their mean anomalies AM. This angular separation can also be easily converted into the length of the circular segment or the time the satellite takes to traverse this segment.

[0067] The radargrammetric data pair, i.e., the radargrammetric survey, is formed by the radargrammetric supplementary satellite and one of the interferometric satellites. Since two interferometric satellites are available, it is also possible to form two radargrammetric data pairs, Al-Bl / A2-B1, for example, to improve the signal-to-noise ratio (by averaging the coordinates of the same object measured in both pairs). The two satellites flying in close formation provide interferometric data pairs, while each of these two satellites, together with one or both of the supplementary satellites, provides radargrammetric data pairs.

[0068] Alternatively, interferometric measurements can also be performed from a single orbital platform if the two antennas for the interferometric measurement are separated by a larger structure that, in at least one dimension, has a size that is substantially equal to or greater than the baseline length required for interferometry. This would be possible, for example, on the International Space Station (ISS).

[0069] The following describes various three-satellite configurations suitable for determining the geographic coordinates of radar point targets. In these configurations, all satellites fly in near-circular orbits with the same orbital period (same mean orbital altitude).

[0070] Satellite Configuration 1: The interferometry satellites fly in close formation. To establish the effective baseline required for interferometry (on the order of 10 2 m to 10 3To achieve the desired amplitude (m, depending on the frequency band), various formations are possible, realized through differences in the orbital elements. These can be differences in eccentricity and / or argument of the perigee (e.g., the so-called "cartwheel formation" [3]), in the ascending node and / or inclination (e.g., the formation referred to as "cross-track pendulum" in DE 101 32 723 B4), or a combination of the aforementioned differences (e.g., the so-called "helix formation" used in TanDEM-X, or relative eccentricity and inclination vector separation, which is also described in DE 101 32 723 B4). Additionally, an offset in the along-track direction can be present due to a difference in the mean anomaly (see also DE 101 32 723 B4).

[0071] In DE 101 32 723 B4, the cartwheel concept according to [3] (CNES) is transferred to claim 1 (so-called cross-track pendulum) in order to generate a vertical cross-track distance in addition to the horizontal cross-track distance resulting from the cross-track pendulum.

[0072] Similarly, the interferometric formation described here can be created using the cartwheel concept or the satellite configurations described in DE 101 32 723 B4. This interferometric formation, which is considered state-of-the-art, is subsequently extended to include a radargrammetry satellite whose relative geometry cannot be realized using the concepts for the interferometry satellite configurations described above due to the different observation requirements.

[0073] The radargrammetry augmentation satellite must maintain a significantly larger horizontal distance to the interferometric formation (on the order of 10).5m). This is preferably achieved by different ascending nodes (difference of ascending nodes: AQ) and / or inclinations. Example of an X-band SAR satellite configuration (Figs. 2 and 3): Satellite Ai flies on a sun-synchronous reference orbit at an altitude of approximately 500 km. Satellite A2 flies in a helical formation with Ai. Together they form the interferometric formation with a maximum cross-track distance of 300 m: the eccentricity difference is 1.5E⁻⁵ (dimensionless) and the difference in the ascending node is 2.5E⁻³°. The radar grammetry augmentation satellite Bi flies on an orbit whose ascending node differs by AQ = -3.6° with respect to the reference orbit of Ai, resulting in an cross-track distance of 436 km at the node transits.For safety reasons, the Bi satellite can be additionally separated by, for example, 76 km (corresponding to 10 seconds of flight time) in the along-track direction with respect to the interferometric formation A1 / 2. This separation also allows the interferometric and radiometric satellites to operate actively without the need for complex synchronization for bi-static operation.

[0074] Fig. 2 shows a representation of the Earth-fixed ascending orbits of the interferometry satellites Ai, ..., Am with m > 2 (continuous orbit) and of the radargrammetry satellite(s) Bi, ..., B nwith n > 1 (dashed path). The orbital planes are separated by differences in the length of the ascending nodes (difference: AQ). The points on the paths symbolize the satellites at the time of their overflight of Central Europe. The depicted 3-minute-long swath over Paris and London is recorded simultaneously by the interferometry formation and the radargrammetry satellite.

[0075] In Fig. 2, the orbits of satellites Ai and A2 cannot be distinguished due to their small spatial separation. In the enlarged view in Fig. 3, the distance between the interferometric satellite orbits Ai and A2 is shown magnified 100 times for a pass over Paris at 49 degrees latitude. The interferometric satellite pair Ai, A2 observes Paris facing right with steep angles of incidence 0AI = 0A2 = 35°. The radargrammetry complementary satellite observes the same scene at a shallow angle of incidence 0BI = 51.6°. The resulting radargrammetric aperture angle is 16.6°.

[0076] The following diagrams (Figs. 4 to 6) show the geometric parameters resulting from the satellite configuration shown in Figs. 2 and 3 as a function of latitude. The gray dashed areas (20 to 55° latitude) are of particular interest for coordinate determination. The configuration parameters were selected accordingly so that the radargrammetric baseline, radargrammetric aperture angle, and interferometric effective baseline in this latitude range meet the requirements for coordinate determination measurements.

[0077] Satellite configuration 2

[0078] The interferometry formation corresponds to the description of the formation in satellite configuration 1. The radargrammetry satellite flies in a similar orbit to satellites Ai and A2, but is located at a different mean anomaly (difference of mean anomalies: AM), causing it to lead or lag the interferometry satellites by a few minutes of flight time. The stereoscopic baseline then results from the displacement of the observation target due to the Earth's rotation relative to the inertially fixed satellites at the overflight times.

[0079] Example of satellites in near-polar orbits (Fig. 7): The interferometric satellite pair Am crosses the equator in an ascending direction at time tO. Looking to the right, they see the ground target at a steep angle of incidence. Satellite B2 crosses the equator 14.4 minutes later, so that, due to the Earth's rotation, the ground target is 436 km further east and is therefore seen at a shallower angle of incidence.

[0080] In practice, medium angles of incidence are preferred for the InSAR method; for the radargrammetric method, a shallower angle of incidence is advantageous as a second angle of incidence. To achieve this observation geometry even in a descending orbit, left-facing observations can be made.

[0081] The radargrammetry satellite Bi shown in Fig. 7 corresponds to the satellite B shown in Fig. 2. nSatellite B2, which acquires a ground target (here Paris) almost simultaneously with the interferometric formation Am, is located on the same inertial orbit as Ai, but at a different mean anomaly. The difference shown, AM = 54.65°, corresponds to a flight time difference of 14.4 minutes at a mean orbital altitude of 505 km. Due to the Earth's rotation, the radargrammetry satellite B2 then sees exactly the same ground targets with the same geometry as satellite Bi, but 14.4 minutes later.

[0082] As with satellite configuration 1, the radargrammetric baseline shortens with increasing latitude due to the lower surface velocity. In the example chosen here, with a difference in the mean anomaly AM = 54.65°, favorable observation geometries result for latitudes between 20° and 55°. The geometric parameters in Figures 4 to 6 also apply to this example. However, the radargrammetric acquisition takes place 14.4 minutes after the interferometric acquisition.

[0083] Differences compared to satellite configuration 1 are as follows:

[0084] Configuration 2 can be realized from an existing / planned constellation of radar satellites by appropriately selecting the mean anomaly (e.g. as a secondary mission).

[0085] Only one data downlink station is required: the interferometric satellites Ai and A2 can transmit their data simultaneously (due to their very small spatial separation) and satellite B2 can then transmit it to the same ground station.

[0086] The interferometric and radargrammetric measurements are no longer taken simultaneously, but with a delay of a few minutes. Satellite configuration 3

[0087] The described satellite configurations 1 and 2 can also be combined. One could choose a fixed difference in the mean anomaly, for example, to obtain a desired revisit geometry, and additionally set a difference in the ascending node to achieve optimal observing conditions for specific latitudes. Furthermore, in the presented configurations, a difference in the inclination of satellite B could also be incorporated. nto enable applications at high latitudes (disadvantage: increased effort required to control the satellite B's then no longer sun-synchronous orbit) n ).

[0088] The following boundary conditions apply to all satellite configurations:

[0089] If the separation of the third satellite occurs via a constant difference in the mean anomaly and / or in the length of the ascending node, optimal application is only possible for certain latitudes. The resulting different coverage patterns for standard images are advantageous, allowing for coverage of a larger area on the ground or a reduction in revisit time.

[0090] To achieve optimal radar angles of incidence (steep for the interferometry formation and shallow for the radargram satellite), the radar antenna must be switched between left- and right-facing orientations twice per orbit. In the examples described above, acquisitions are performed right-facing on ascending orbits and left-facing on descending orbits. By selecting different configuration parameters, observations can also be performed left-facing on ascending orbits and right-facing on descending orbits.

[0091] If a second satellite flying in close formation is placed alongside the radargrammetry satellite, a second interferometry formation is obtained. With such a satellite configuration, the need to switch between right- and left-facing orientations is eliminated, as explained below using the example of two satellite pairs, A (with satellites Ai and A2) and B (with satellites Bi and B2), which maintain a constant right-facing orientation. Satellite pair Ai and A2 perform interferometric observations in ascending orbits at steep angles of incidence; in descending orbits, the radargrammetric measurements are performed using only satellite Ai or A2 at a shallow angle of incidence. Accordingly, radargrammetric measurements are performed with Bi or B2 during ascending orbits, and interferometric measurements are performed with Bi and B2 during descending orbits.

[0092] If additional radargrammetry satellites or satellite pairs become available, further observation pairs can be formed to achieve, for example, a favorable observation geometry over a wider range of latitudes and to increase the number of ground-based objects that can be detected and whose coordinates can be measured. For example, another satellite, Ci, could have a radargrammetric baseline of 300 km to the interferometry formation during its equator crossing. This would allow areas below 20 degrees latitude to be surveyed as well.

[0093] Processing of the data acquired with the SAR satellite information according to the invention for the precise determination of geo-coordinates on the Earth's surface

[0094] Fig. 8 shows an overview of the data processing when using the SAR satellite configuration according to the invention for determining the geo-coordinates of radar point targets. The illustration refers to the application case with two interferometric and one radargrammetric satellite (when processing a radargrammetric pair), but is transferable to all combinations of at least two first interferometric and at least one second radargrammetric satellite. On the input side, the SAR images of the interferometric satellite pair, designated Ai and A2, as well as the SAR image of the more distant satellite enabling radargrammetry, designated Bi, are available.

[0095] The image coordinates of the SAR data Ai, A2 and Bi are corrected by geodetic correction for tropospheric, ionospheric and for errors caused by Earth tides in the range and azimuth directions (geodetic correction, GECO), resulting in the SAR data AI_GECO, A2_GECO and BI GECO annotated with corrected coordinates.

[0096] In the image data BI_GECO and AI_GECO, a point detection takes place, from which the 2-D point sets PBI and PAI are derived.

[0097] Using stereo-SAR methods, the point sets PAI and PBI are assigned to each other, where possible, and combined to form GCPAiBi_stereo, a set of 3-D Ground Control Points (GCP) or estimates of their 3-D accuracy.

[0098] In a point selection, the subset of the most accurate points, GCPcaiibration, is extracted from the set GCPAiBi_stereo.

[0099] The geodetic-corrected data of the interferometric part, AI_GECO and A2_GECO, are corrected by co-registration, i.e., by estimating the position-dependent mutual displacements A rg and A az The range and azimuth are mapped onto each other, and then A2_GECO is resampled based on this. The result is A2_GEco_Res, an image geometrically matching AI_GECO.

[0100] The phase values ​​of the complex-valued images AI_GECO and A2_GEco_Res are combined in interferogram formation to form the interferometric phase ([•wrapped]), which is ambiguous. Phase unwrapping transforms t|> wrapped into the interferometric phase ([»unwrapped]), which is free of ambiguities.

[0101] The unwrapped phase is height-dependent, but overall (as an image), when converted to height values, it exhibits an offset from actual heights. Furthermore, each individual point in the unwrapped phase is subject to an individual error resulting from the geodetically influenced range measurements. Phase calibration compensates for both the offset and the pixel-specific geodetic phase errors. The offset is determined by comparing the unwrapped phase with the height values ​​of the high-precision 3D points (GCP calibration) of the stereo-SAR (radargrammetry) branch, which are converted to phases using H-<|) conversion. Pixel-specific phase errors are corrected using the differential geodetic range errors (ARA2-AI_GECO = ARA2_GECO - ARAI_GECO), which are converted to phase values ​​in the GECO phase correction. The result is the calibrated unwrapped phase (<|).

[0102] The phase <|)unwrapped_caiib is converted into a very accurate digital terrain model (DEM) during DEM generation.

[0103] Based on the generated, highly accurate DEM, the set of points detected in the BI GECO image is directly georeferenced in the geocoding step. The result is a denser network of 3D geographic coordinates, whose accuracy surpasses that of conventional methods.

[0104] The quantities GCPBI_GEOCODED and GCPCALIBRATION are combined to form the final product.

[0105] Further embodiments of the invention

[0106] The "Mirror SAR Principle" [4] could be used, which simplifies SAR satellites when they fly in formation. Operators of large SAR satellite formations could use the concepts disclosed here for their satellite orbits and thus gain additional functions for their formation, such as the ability to determine the position of objects on the ground, without additional hardware costs.

[0107] To enable two-dimensional data acquisition when measuring surface currents at sea or when measuring flowing traffic (Ground Moving Target Indication, GMTI), different viewing angles of the area on the ground are also necessary. This could be achieved by a close formation of SAR satellites that illuminate the scene from different squint angles. However, operators of large satellite formations prefer to distribute their satellites as widely as possible in different orbits to obtain a shorter revisit time for a region on the ground. The satellite configurations according to the invention provide ways to simultaneously meet both requirements (distributed satellites and observation of an area from different viewing angles).

[0108] The situation is very similar with the so-called "SAR Theater Mode", in which several images are desired in an area, but these cannot be achieved simultaneously with one satellite.

[0109] List of abbreviations

[0110] DGM Digital Terrain Model GECO Geodetic Correction GCP Ground Control Point GMTI Ground Moving Target Indication GNSS Global Navigation Satellite System

[0111] PSI Persistent Scatterer Interferometry SAR. Synthetic aperture radar SCR signal-to-clutter ratio

[0112] Bibliography

[0113] W02011 / 092056 A2

[0114] DE 101 32 723 B4

[0115] DE 10 2010 001 440 Al

[0116] DE 10 2016 208 508 Al

[0117] [1] Raggam, H., Gutjahr, K., Perko, R., Schardt, M., "Assessment of the Stereo-Radargrammetric Mapping Potential of TerraSAR-X multibeam spotlight Data", IEEE Transactions on Geoscience and Remote Sensing, 48(2), pp. 971- 977, 2010. https : / / ieeexplore.ieee.org / stamp / stamp.jsp?arnumber=5361396

[0118] [2] Gisinger, C., Libert, L., Marinkovic, P., Krieger, L., Larsen, Y., Valentino, A., Breit, He., Balss, U., Suchandt, S., Nagler, T., Eineder, M., Miranda, Nuno, The Extended Timing Annotation Dataset for Sentinel-1— Product Description and First Evaluation Results, in IEEE Transactions on Geoscience and Remote Sensing, vol. 60, pp. 1-22, 2022, Art no. 5232622, doi :

[0119] 10.1109 / TGRS.2022.3194216.

[0120] [3] Massonnet, D. (CNES), Capabilities and Limitations of the Interferometric Cartwheel, in Proceedings of the CEOS SAR Workshop, Toulouse, 26-29 October 1999, ESA SP-450, March 2000.

[0121] [4] Krieger, G., Zonno, M., Rodriguez Cassola, M., Lopez Dekker, Paco., Mittermayer, J., Younis, M., Huber, S., Villano, M., Queiroz de Almeida, F., Prats, P., Moreira, A. : MirrorSAR, A Fractionated Space Radar for Bistatic, Multistatic and High-Resolution Wide-Swath SAR Imaging. In: International Geoscience and Remote Sensing Symposium (IGARSS). IEEE. IEEE International Geoscience and Remote Sensing Symposium (IGARSS), 2017- 07-23 - 2017-07-28, Fort Worth, USA, doi : 10.1109 / IGARSS.2017.8126916. https: / / elib.dlr.de / 114778 /

Claims

REQUIREMENTS 1. Satellite configuration for remote sensing of the Earth's surface, in particular for near real-time determination of the geo-coordinates of radar point targets, comprising at least two first SAR satellites and at least one second SAR satellite, wherein the at least two first SAR satellites are intended for interferometric surveying of a survey area on the Earth's surface and each pair of the first SAR satellites has an cross-track separation of 10 m to 1 km, wherein the at least one second SAR satellite establishes a radar-grammetrically effective cross-track separation of 100 km to 1000 km with each of the first SAR satellites, and wherein the at least one second SAR satellite is intended for radar-grammetric surveying of the survey area surveyed by the at least two first SAR satellites.wherein the interferometric measurements can be calibrated using the radargrammetric measurements, wherein all first and second SAR satellites have the same illumination direction with respect to the orbital direction, and wherein the illumination direction of all first and second satellites is simultaneously either left-looking or right-looking.

2. Satellite configuration according to claim 1, characterized in that the at least two first SAR satellites move at the same mean speed in different, nearly circular orbits at the same mean altitude, wherein the orbits differ in the orbital elements eccentricity and / or argument of perigee, thereby creating a vertical cross-track distance between the satellites, which increases in length over one orbital revolution. periodically varied, so that areas of interest on Earth can be interferometrically measured with a sufficiently effective baseline.

3. Satellite configuration according to claim 1, characterized in that the at least two first SAR satellites move at the same mean speed in different, nearly circular orbits at the same mean altitude, wherein the orbits lie in different planes, i.e., have different lengths of the ascending node and / or different inclinations, thereby creating a horizontal cross-track distance between the satellites, the length of which varies periodically over one orbital revolution, so that areas of interest on Earth can be interferometrically measured with a sufficiently effective baseline.

4. Satellite configuration according to claim 1, characterized in that the at least two first SAR satellites have a vertical and horizontal cross-track offset with the same or different amplitude, so that areas of interest on Earth are interferometrically measured with a sufficiently effective baseline.

5. Satellite configuration according to claim 1, characterized in that the at least two first SAR satellites have, in addition to a vertical and / or horizontal Across-Track offset, an Along-Track offset caused by different mean anomalies of the two orbits.

6. Satellite configuration according to claim 1 and according to any one of claims 2 to 5, characterized in that the at least one second SAR satellite moves with the same mean speed relative to the at least two first SAR satellites on a different, nearly circular orbit at the same mean altitude, wherein the orbits lie in different planes, i.e., different lengths of the ascending node and / or different inclinations exhibiting a horizontal cross-track distance between the first and second SAR satellites, which varies periodically in length over one orbital revolution, so that areas of interest on Earth are measured with a sufficiently radargrammetrically effective cross-track baseline.

7. Satellite configuration according to claim 1 and according to one of claims 2 to 5, characterized in that the at least one second SAR satellite has, in addition to the horizontal Across-Track distance, an Along-Track offset with respect to the at least two first SAR satellites, which is caused by different mean anomalies.

8. Satellite configuration according to claim 1 and according to one of claims 2 to 5, characterized in that the at least one second SAR satellite has, in addition to the horizontal cross-track distance, a vertical cross-track offset with respect to the at least two first SAR satellites, which is caused by differences in the orbital elements eccentricity and / or argument of perigee.

9. Satellite configuration according to claim 1 and according to one of claims 2 to 5, characterized in that the at least one second SAR satellite leads or lags behind the at least two first SAR satellites, which is caused by different mean anomalies, so that the radargrammetrically effective Across-Track-Baseline follows from the displacement of the observation target due to the Earth's rotation between the overflight times of the first and second SAR satellites.

10. Satellite configuration according to any one of claims 2 to 9, characterized by at least one further second SAR satellite which provides an Across-Track Baseline of 10 to each of the other second SAR satellites 1 m to 10 3 m, each of which has an Across-Track baseline of 10 with each of the first SAR satellites 5 m to 10 6 m forms, with the first SAR satellites and the second SAR satellites per The orbits change their functions twice: in the ascending orbit, the first SAR satellites are used for interferometry surveying, and at least one of the second SAR satellites, or one of the second SAR satellites together with at least one of the first SAR satellites, is used for radargrammetry surveying; and in the descending orbit, the second SAR satellites are used for interferometry surveying, and at least one of the first SAR satellites, or at least one of the first SAR satellites together with at least one of the second SAR satellites, is used for radargrammetry surveying.

11. Method for determining the geo-coordinates of radar point targets, such as radar point spreaders, wherein in the method a satellite configuration according to one of claims 1 to 10 is provided, an interferometry survey of a survey area on the Earth's surface is carried out using the first SAR satellites, a radargrammetry survey of the survey area is carried out using at least one second SAR satellite together with one of the first SAR satellites, for which an interferometry survey has been carried out using the first SAR satellites, and the measured values ​​of the interferometry survey are calibrated using the measured values ​​of the radargrammetry survey.

12. Method according to claim 11, characterized in that the satellite configuration is supplemented by at least one further second SAR satellite which provides an Across-Track Baseline of 10 to each of the other second SAR satellites 1 m to 10 3 m, each of which has an Across-Track baseline of 10 with each of the first SAR satellites 5 m to 10 6 m forms, and that the first SAR satellites and the second SAR satellites per orbit change their functions twice, with the first SAR satellites becoming the second SAR satellites in the ascending orbit. Interferometry surveying is planned, and at least one of the second SAR satellites or one of the second SAR satellites together with at least one of the first SAR satellites or with at least one other of the second SAR satellites is planned for radargrammetry surveying, and in the descending orbit the second SAR satellites are planned for interferometry surveying, and at least one of the first SAR satellites or at least one of the first SAR satellites together with at least one of the second SAR satellites or with at least one other of the second SAR satellites is planned for radargrammetry surveying.

13. Method for processing the images from SAR satellites of a satellite configuration comprising a first and a second interferometry SAR satellite and at least one radargrammetry SAR augmentation satellite for forming the geo-coordinates of radar point targets on the Earth's surface, wherein in the method a) interferometry images with the 2D coordinate grids Ai, A2 are provided by the two interferometry SAR satellites and a radargrammetry image with the 2D coordinate grid Bi is provided by the at least one radargrammetry SAR augmentation satellite, b) the image coordinate grids of the first interferometry SAR satellite are geodetic corrected to form the first corrected interferometry image coordinate grid AI_GECO,c) the image coordinate grids of the second interferometry SAR satellite are processed by geodetic correction to form a second corrected interferometry image coordinate grid A2_GECO and the image coordinate grid of at least one radar grammetry SAR supplementary satellite are processed by geodetic correction to form a corrected radar grammetry image coordinate grid BI_GECO, c) objects with point-like radar backscatter are detected in the SAR image of the first interferometry SAR satellite and their coordinates with respect to the first corrected interferometry, a) determine the image coordinate grid AI_GECO, which forms a first 2D point set PAI_GECO; b) detect objects with point-like radar backscatter in the SAR image of the radar grammetry SAR augmentation satellite and determine their coordinates with respect to the corrected radar grammetry image coordinate grid BI_GECO, which form a second 2D point set PBI_GECO; c) determine the point pair set MAIBI_GECO of those pairs of points that are contained in the first and second 2D point sets PAI_GECO and PBI_GECO and whose mutual distance is below a specified threshold based on the distance between the image coordinates of the points from the first 2D point set PAI_GECO and those of the points from the second 2D point set PBI_GECO.f) using a stereo-SAR method based on the parallax of their image coordinates, the elements of the point pair set MAIBI_GECO are combined to form a set of 3D coordinates GCPAiBi_stereo of the point targets with an estimated accuracy, g) from the set of 3D coordinates GCPAiBi_stereo, those coordinates that have an accuracy better than a predefinable threshold are combined to form a subset GCPcaiibration of the point targets with the most accurate 3D coordinates, h) the two geodetically corrected interferometry image coordinate grids AI GECO, AZ GECO are combined by estimating their position-dependent displacements relative to each other in Range A, rgand azimuth Aaz are mapped onto each other, i) then, based on this mapping, the second corrected interferometry image coordinate grid A2_GECO is resampled to form a second corrected interferometry image coordinate grid A2_GEco_Res that is geometrically congruent (co-registered and re-sampled) to the first corrected interferometry image coordinate grid AI_GECO, j) the phase values ​​of the complex-valued interferometry images in the first and in the co-registered second corrected interferometry image coordinate grid AI_GECO, A2_GECO_RES are combined by interferogram formation to form the ambiguous interferometric phase <|) wrapped, k) the interferometric phase <|) unwrapped, freed from ambiguity, is formed by phase unwrapping, l) by phase calibration both a possible offset of the interferometric phase ^unwrapped to the height values ​​of the points GCPcaiibration converted into phase values ​​(and thus to actual heights) and, by using the GECO phase difference A<|)GECO formed from the geodetic corrections ARAI_GECO and ARA2_GECO converted into phases, image position-dependent, individual geodetic phase errors are compensated and the calibrated phase <|) unwrapped_calib is obtained,m) the calibrated phase <|) unwrapped_calib is converted into a digital elevation model (DEM) using a method for creating a digital elevation model (DEM), n) all of the radar grammetry image points PBI_GECO not used in step g) in the set of geodetically corrected detected radar grammetry image points are directly georeferenced to a set of 3D GCP coordinates GCPBI_GEOCODED using a geocoding method based on the generated digital elevation model, and o) the sets GCPBI_GEOCODED and GCPCALIBRATION are combined to form the final product.

Images