The SRTM mission history: the NASA SIR series and the Italian and German
experiences on space radar missions as precursors of the present mission

NASA/JPL's radar program began with Sea Surveillance Satellite (SEASAT), the first spaceborne imaging radar, in 1978. SEASAT was a single frequency (L-band; l ~ 24 cm), single polarization, fixed-look angle (20^o) radar, launched into a 108^o inclination and an 800 km altitude orbit. The purpose of the mission was oceanographic studies, but the radar images were also useful for geological science. The mission ended after 100 days (due to a power system failure), but many images were still returned. Approximately 10⁸ sq. km were mapped.

The next four radar Earth-mapping missions were SIR-A on STS-2 (11/81), SIR-B on STS-41G (10/84), and SRL-1 and SRL-2 (a combination of SIR-C and X-SAR) on STS-59 and STS-68, respectively (4/94 and 10/94).

SIR-A was an L-band radar with a fixed look angle of 50^o. All SIR-A data were optically recorded and correlated. SIR-A imaged buried drainage channels in southern Egypt, cloud covered regions, and dynamic ocean phenomenon. About 10⁷ sq. km were imaged with a 50 km swath width and a 40 m resolution.

SIR-B added a variable look-angle capability and digital image recorder to the L-band, single polarization radar. Images were recorded and downlinked through the 46 Mb/s Tracking and Data Relay Satellite System (TDRSS) Ku channels. Because of problems with the Ku system, the swath width varied from 10 km to 50 km. SIR-B generated the first stereo radar images, which were used to create 3-dimensional topographic maps. The variable incidence angle was used to identify radar signatures of different terrain and vegetation.

SIR-C provided increased capability acquiring digital images simultaneously at two microwave wavelengths, L- band and C-band. These vertically and horizontally-polarized transmitted waves were received on two separate channels, so that SIR-C provided images of the magnitude of radar backscatter for four polarization combinations: HH (Horizontally-transmitted, Horizontally-received), VV (Vertically-transmitted, Vertically-received), HV, and VH; and also data on the relative phase difference between the HH, VV, VH, and HV returns. This allowed derivation of the complete scattering matrix of a scene on a pixel-by-pixel basis.

Germany's imaging radar program started with the Microwave Remote Sensing Experiment (MRSE) flown aboard the first Spacelab mission in 1983. The program was continued by development of the X-SAR, for which cooperation with Italy was initiated.

The combined SIR-C/X-SAR system, otherwise known as SRL, provided a capability to acquire images at three microwave wavelengths, L-band (l ~ 24 cm) quad-polarization; C-band (l ~ 6 cm) quad- polarization; and X-band (l ~ 3 cm) single (VV) polarization. Because radar backscatter is most strongly influenced by objects comparable in size to the radar wavelength, this multifrequency capability provided information about the Earth's surface over a wide range of scales not discernible with previous single-wavelength experiments. SIR-C/X-SAR also had variable look angles, and could image at incidence angles between 20^o and 65^o. Typical image sizes for SIR-C data products were 50 km x 100 km, with resolution between 10 m and 25 m in both dimensions.

In addition to the imaging modes required to accomplish the scientific objectives of the SRL flights, a number of experimental modes were incorporated into the SIR-C design for technology development. Among these was SCANSAR, a technique for imaging across a very wide swath by electronically sweeping the radar beam in the cross-track direction (perpendicular to the line of flight).

Using SCANSAR, SIR-C was able to image at both C-band and L-band across a 250 km swath. It is this wide swath that makes the SRTM mission possible.

One of the major accomplishments of the SRL flights was the demonstration of "pass-to-pass" interferometry, a technique for measuring the height of topography (or very small changes in that height). Figure 1 shows how data from two imaging passes can be combined. The fundamental measurement produced by any SAR is the range from the radar antenna to points on the ground, denoted for example by S₁ in the figure. S₁’ is the range to that same point measured on a different imaging pass. Since the measurements are accurate to a fraction of a wavelength, the difference S₁ - S₁’ for SIR-C or X-SAR can be determined to a fraction of a centimeter. Of course, this so-called phase difference is modulated at each point by the height of the topography. Given the spacecraft positions, the phase difference for the case of perfectly smooth terrain can be predicted mathematically, any residual present after this "smooth earth" function is subtracted is directly proportional to the topographic height.

Fig. 1: Geometry for pass-to-pass interferometry

During the last three days of the SRL-2 mission, Flight Dynamics Officers (FDOs) were successful in repeating the shuttle’s orbital path from day-to-day almost exactly, necessary to minimize the interferometric baseline. Baselines greater than about a kilometer would result in data that do not correlate well enough for interferometry, but baselines of less than a few hundred meters were achieved and numerous high quality topographic maps were generated using data from all three SIR-C/X-SAR wavelengths.

Instead of pass-to-pass interferometry, SRTM will use a 60-meter long space station-derived mast to perform fixed-baseline interferometry. This fixed-baseline interferometry uses two separated antennae simultaneously with one antenna transmitting radar pulses and both antennae receiving echoes, so that only one data pass is required for each data swath. The transmitting antennae will be located within the payload bay, as they were for the SRL missions, and the receive-only antennae will be located at the end of the mast. This technique has the advantages of avoiding temporal decorrelation caused by changes in the terrain between imaging passes, of keeping the baseline separation between antennae constant, and of keeping the antenna orientation well-determined.