This system could be used to map earthquakes, landslides, floods, and wildfires.
A computer system denoted a change-detection onboard processor (CDOP) is being developed as a means of processing the digitized output of a synthetic-aperture radar (SAR) apparatus aboard an aircraft or spacecraft to generate images showing changes that have occurred in the terrain below between repeat passes of the aircraft or spacecraft over the terrain. When fully developed, the CDOP is intended to be capable of generating SAR images and/or SAR differential interferograms in nearly real time. The CDOP is expected to be especially useful for understanding some large-scale natural phenomena and/or mitigating natural hazards: For example, it could be used for near-real-time observation of surface changes caused by floods, landslides, forest fires, volcanic eruptions, earthquakes, glaciers, and sea ice movements. It could also be used to observe such longer-term surface changes as those associated with growth of vegetation (relevant to estimation of wildfire fuel loads).
The CDOP is, essentially, an interferometric SAR processor designed to operate aboard a radar platform. The CDOP design features a compact processor architecture chosen to combine the flexibility of microprocessors with the very high speed of field-programmable gate arrays (FPGAs) so as to optimize throughput performance while maintaining flexibility. The processor design addresses three critical requirements of real-time, onboard processing hardware for interferometric SAR imaging: high computational throughput, a large amount of onboard memory, and high-speed data interconnections throughout the processing chain. The functional blocks of the CDOP (see figure) include the following:
- Preprocessor — A microprocessor within a control-and-interface computer serves as a preprocessor that generates parameters necessary for generation of SAR image data from a combination of ephemeris and radar-configuration data. For the ephemeris data, the preprocessor implements a six-state (three position and three velocity coordinates) Kalman filter to effect real-time reconstruction of the platform trajectory from the outputs of an inertial navigation unit and a Global Positioning System receiver. The preprocessor also includes an azimuth pre-summer to decimate and re-align range-compressed data in the along-track direction, and an optional motion compensation- module for airborne interferometric SAR. The output of the preprocessor is utilized by the SAR image formers described next.
- SAR Image Formers — The raw SAR data are processed into SAR image data by two FPGA SAR processors that implement a range-Doppler algorithm with motion-compensation capability. The two SAR processors accept two input streams of raw radar data: typically, these would be (1) the real-time stream of data from the high-speed digital back end of the operating radar apparatus and (2) a stream of corresponding data from a previous pass retrieved from a high-speed storage device via a fiber-optic link. Alternatively, if the radar platform were to include two radar apparatuses in an interferometric configuration, then the CDOP could process the near-real-time streams of data from these apparatuses for use in generating a single-pass interferogram.
- Interferometric Postprocessor — Another microprocessor generates SAR interferometric image data from the outputs of the two SAR processors. In a typical application, the output of this processor would be transmitted to a ground station (downlinked). Because all of the processing up to the point of downlinking would be done onboard, the downlink data rate necessary for observing changes in the terrain would be significantly reduced.
This work was done by Yunling Lou, Ronald J. Muellerschoen, Steve A. Chien, and Sasan S. Saatchi Caltech and Duane Clark of Leeward Engineering for NASA’s Jet Propulsion Laboratory.
This Brief includes a Technical Support Package (TSP).
Onboard Data Processor for Change- Detection Radar Imaging (reference NPO-45751) is currently available for download from the TSP library.
Please Login at the top of the page to download.