TerraSAR-X and TanDEM-X 'chatting' for the first time
Even though the 'chit chat' between the satellites has so far consisted of no more than the exchange of synchronisation signals – "Hello, here I am!" – that is been all that was needed for the first simultaneous imagery to be captured by TerraSAR-X and TanDEM-X. Since 22 July 2010, TerraSAR-X has been flying ahead of TanDEM-X at a distance of just 20 kilometres, in a formation defined for test purposes and which also affords scope for a special series of radar experiments. They have now completed the first ever experiment with bistatic radar involving two satellites flying together in formation.
What does 'bistatic radar' mean?
The phrase 'bistatic radar' refers to the transmission and reception antennae on a radar system being at different locations. In a classic radar configuration, the same antenna is used to transmit and to receive, and this kind of system is also referred to as 'monostatic' radar.
The bistatic formation
In our experiment, TerraSAR-X was the active part of the bistatic configuration and its radar was operated conventionally; that is, monostatically. That means that its antenna is used for transmission and also for reception (red arrows in the figure). TanDEM-X, flying behind the other satellite, is the passive part of this configuration because its antenna is only used to receive the signals transmitted by TerraSAR-X and reflected back off the Earth’s surface (green arrows in the figure). The operating mode employed by TanDEM-X is also referred to as 'parasitic operation' because it does not transmit, and instead only 'listens in to' echoes generated in response to transmissions by the other radar. Working together in this way, the satellites achieve 'bistatic' radar operation.
However, the distance of 20 kilometres separating the satellites is too great for both radars to be able to 'see' the area being recorded at the same time during normal operation. As a consequence, the directions of view of the two radars need to be calibrated accordingly and deliberately 'misaimed' at the recording area located centrally beneath the orbital paths of the two satellites. This redirection enables the antenna arrays on TerraSAR-X and TanDEM-X to activate their 384 transmit/receive modules electronically in such a way that the radar beam alters its direction of view. That avoids the need for any mechanical rotation of the satellites. To produce successful bistatic imagery, the experts at the DLR Microwaves and Radar Institute arranged for TerraSAR-X, flying at the front of the formation, to be 'looking' about 0.8 degrees behind itself, and for the TanDEM-X, flying behind its counterpart, to be 'looking' about 0.8 degrees ahead of itself.
An additional requirement is for both radar systems to operate synchronously. That means that the TanDEM-X satellite, which only receives signals, needs to ‘know’ when to expect an echo and then to open and close its reception aperture at the correct times. It accomplishes this feat by analysing GPS time and the exchange of synchronisation signals using highly-specialised additional antenna arrays located on both satellites.
What is expected from this?
Two SAR images are obtained from the formation described here. The image from TerraSAR-X differs only slightly from the conventional SAR images, which are now familiar from TerraSAR-X operation prior to this experiment. The image is, of course, generated in 'classic' or 'monostatic' mode and is based on the radar echoes reflected back by the surface of the Earth. The second and more interesting image is based on signals received by TanDEM-X. This is in part created from the echoes reflected away from TerraSAR-X and which no single satellite operating alone would be able to receive. In other words, additional information is obtained about the area being imaged; information that TerraSAR-X would normally be incapable of receiving and which would otherwise be missed altogether.
Quite apart from its broader scientific aspects, this imagery enables us to complete our scheduled test programme. This programme includes validation of the synchronisation technique described above, which plays a significant role in generation of the digital terrain models that are the primary objective of this mission.
The experiment and its first results
The experiment was conducted over Brasília, the capital of Brazil, which can be seen to the left of the image and which shows the characteristic crucifix outline of this city. This image is obtained by superimposing the bistatic image, shaded in green (acquired by TanDEM-X) and the simultaneous monostatic image, shaded in magenta (acquired by TerraSAR-X).
The next image shows an enlargement of the inner city area and employs the same colour coding. The distribution of these colours highlights the difference in the sensitivity of the surface for the monostatic and bistatic reflection processes; that is, the forward and back scattering of the signals. Despite the very slight difference in the angle of view of the two satellites, differences stand out clearly.
Inner city area of Brasília with its characteristic outline. When the bistatic image (green) and the monostatic image (magenta) are superimposed, the latter dominates in this section of the resultant picture.
However, for future applications, the underlying scatter mechanisms need to be studied in greater detail. It is nevertheless becoming apparent that the combination of monostatic and bistatic modes can contribute substantially towards the classification of land surfaces.
Sobradinho to the north of Brasília. Superimposition of the bistatic (green) and the monostatic image (magenta) indicates a dominant bistatic component in the depiction of this residential area.
I would not like to end this blog entry without first extending my grateful thanks to my colleagues from the DLR Microwaves and Radar Institute. Without their expertise and their involvement at the planning stage, during the difficult commanding and calibration operations with the radar instruments and adaptation of the TAXI (experimental TanDEM-X Interferometric Processor) processor software developed at the institute, these results could not have been obtained.
Image credits: DLR.