A radar image is the end product of a chain of measurements, assumptions and processing decisions. The deeper scientific opportunity begins when we inspect the evidence that sits behind the display.
Synthetic aperture radar combines transmitted pulses with the motion of the spacecraft to focus echoes into an image. Depending on the product, each pixel can carry amplitude, phase, polarisation and geometric context. Repeat observations add time; different wavelengths and look directions add constraints. None is a universal key. Each dimension answers a different part of the physical question.
Scattering geometry
Where did the return originate, and how do incidence angle, terrain slope, shadow, layover and pixel footprint shape what is represented?
Explore SAR image formation ↗Path length & motion
For coherent complex data, phase differences can encode path changes. Baseline geometry and atmospheric delay must be separated from displacement.
Explore interferometry ↗Repeated observation
A time series can expose persistent trends and seasonal patterns, while temporal decorrelation and irregular acquisitions limit comparison.
Explore research questions ↗Wavelength & polarisation
Microwave wavelength and polarisation influence which structures contribute to the return; they alter sensitivity, not just image appearance.
Explore sensor complementarity ↗Multiple baselines
Different viewing geometries can constrain vertical scattering distributions in some scenes. Tomographic resolution depends on baseline spread and signal coherence.
Explore radar tomography ↗Models & inference
Physical forward models and inverse methods compare candidate explanations against measured data, with priors and uncertainty made explicit.
Explore reconstruction ↗One measurement chain, many opportunities
Calibration establishes how digital values relate to the sensor measurement. Focusing uses the phase history of echoes to form the synthetic aperture image. Geometric and radiometric correction make spatial and intensity comparisons more meaningful. Derived products—interferograms, deformation estimates, land-cover classifications or visualisations—add analytical steps and assumptions. A research workflow should keep the source product, processing lineage and evidence for each conclusion connected.
ViewPiece's technical programme is exploratory. It investigates where improved calibration, phase-aware processing, time-series integration, physical modelling or uncertainty-aware reconstruction could improve a defined analysis. No universal increase in resolution or demonstrated performance is claimed.
Technology research areas
Radar augmentation
Research into processing methods that may make more information in existing radar observations available for analysis.
Question to scope
Which signal properties are being lost, hidden or made difficult to interpret by a given processing chain?
Interferometry
Using phase differences between coherent radar acquisitions to investigate relative changes in the sensor-to-target path.
Question to scope
Are the observations sufficiently coherent and well matched in geometry and time for a meaningful comparison?
Radar tomography
Studying whether multiple coherent observations can distinguish the vertical distribution of radar scattering within a resolution cell.
Question to scope
Do the available baselines and acquisition geometry support the structural question at the scale required?
Computational reconstruction
Exploring inverse methods that estimate scene properties from measurements, with assumptions and uncertainty made explicit.
Question to scope
What scene properties can the available measurements constrain, and which must come from prior assumptions?
Multisensor fusion
Assessing whether radar, optical, elevation and contextual sources can answer a question more reliably together.
Question to scope
Can the datasets be aligned in space, time, scale and reference without concealing disagreement?
Scientific visualisation
Designing views that distinguish input measurements, derived estimates, missing information and uncertainty.
Question to scope
Can an institutional user see what was observed, what was calculated and how confident the interpretation is?
Data architecture
Research into data handling that keeps measurements, metadata and processing history connected throughout analysis.
Question to scope
Can a result be reproduced from traceable inputs without avoidable conversion or loss of relevant information?
Validation & uncertainty
Planning how derived outputs could be compared with independent observations and evaluated for a defined use.
Question to scope
What reference data, metrics, operating conditions and failure cases would make a result meaningful?