4D topography from space combining short-baseline stereo and lidar
Résumé
A new space borne instrumental concept called Z-Earth has undergone phase 0 study at CNES, the French space agency to fulfill these aims. The scientific communities backing this proposal are that of natural hazards, glaciology and biomass. The system under study combines a short-baseline native stereo imager and a lidar profiler. This combination provides spatially resolved elevation swaths together with absolute along-track elevation control point profiles. Acquisition is designed for revisit time better than a year. Intended products not only target single pass digital surface models, color orthoimages and small footprint full-wave-form lidar profiles to update existing topographic coverage, but also time series of them. 3D change detection targets centimetre-scale horizontal precision and metric vertical precision, in complement of -now traditional- spectral change detection. To assess the actual concept value, two real-size experiments were carried out. We used sub-meter-scale Pleiades panchromatic stereo-images to generate digital surface models and check them against dense airborne lidar coverages, one heliborne set purposely flown in Corsica (50-100pts/sq.m) and a second one retrieved from OpenTopography.org (~10pts/sq.m.). In Corsica, over a challenging 45-degree-grade tree-covered mountain side, the Pleiades 2-m-grid-posting digital surface model described the topography with a median error of -4.75m +/-2.59m (NMAD). A planimetric bias between both datasets was found to be about 7m to the South. This planimetric misregistration, though well within Pleiades specifications, partly explains the dramatic effect on elevation difference. In the Redmond area (eastern Oregon), a very gentle desert landscape, elevation differences also contained a vertical median bias of -4.02m+/-1.22m (NMAD). Though here, sub-pixel planimetric registration between stereo DSM and lidar coverage was enforced. This real-size experiment hints that sub-meter accuracy for 2-m-grid-posting DSM is an achievable goal when combining stereoimaging and lidar. The advent of free-of-charge global topographic data sets SRTM and Aster GDEM have enabled testing a host of geoscience hypotheses. This is because they revealed for the first time the relief of previously undescribed earth landscapes, enabled quantitative geomorphometric analyses across entire landscapes and improved the resolution of measurements. Availability of such data is now considered standard, and though resolved at 30-m to 90-m pixel size, which is amazing seeing where we come from, they are now regarded as mostly obsolete given the sub-meter imagery coming through web services like Google Earth. Two additional features would help meet the current data needs of the Geoscience communities:: field-scale-compatible elevation datasets (i.e. meter-scale digital models and sub-meter elevation precision) and provision of regularly updated topography to retrieve earth surface changes, while retaining the key for success: data availability at no charge. A new satellite instrument called Z-Earth has undergone phase 0 study at CNES, the French space agency to fulfil these aims. The scientific community backing this demand is that of natural hazards, glaciology and the biomass community. The system under study combines a native stereo imager and a lidar profiler. This combination provides spatially resolved elevation swaths together with absolute along-track elevation control point profiles. Data generated through this system, designed for revisit time better than a year, is intended to produce not only single acquisition digital surface models, colour orthoimages and small footprint full-wave-form lidar profiles to update existing topographic coverage, but also time series of them of them. This enables 3D change detection fields with centimetre-scale planimetric precision and metric vertical precision, in support of now traditional spectral change detection. Designed to meet the data requirements of natural hazard and glaciology communities, Z-Earth design supports other applications such as forest ecosystems monitoring and biomass assessment. The purpose of this contribution is to present the mission concept and its philosophy and the scientific needs, including foreseen societal benefits, that justify the system development.