Ziyuan: separate related names and map the terrain

Ziyuan, often abbreviated ZY, denotes a Chinese Earth-resources satellite lineage with several branches. Its early ZY-1 history overlaps the China–Brazil Earth Resources Satellite program, CBERS. CNSA’s historical account connects the 1988 bilateral agreement to the spacecraft launched in October 1999, while INPE documents the same shared program. The early spacecraft belong to a shared cooperation history, while the broader resource-observation lineage also includes distinct Chinese missions.

Later names require more care. CNSA describes Ziyuan-1 02D, launched in September 2019, as providing multispectral and hyperspectral data for resource and environmental applications. A separate historical review distinguishes three Chinese Ziyuan-2 spacecraft from two joint China–Brazil satellites. Ziyuan-3 is a stereo-mapping branch. Preserve those distinctions instead of labeling every Ziyuan record CBERS or transferring one spacecraft’s sensor specifications to another.

Stereo mapping compares observations of the same terrain from different directions. The apparent displacement of matching features, together with camera geometry and spacecraft position, constrains surface elevation. CNSA’s December 2025 Ziyuan-3 04 report identifies stereo and multispectral cameras plus a laser altimeter. Multiple sensors provide complementary constraints; they do not make a terrain model independent of calibration and matching errors.

Try a simple geometric analogy: hold a finger in front of a distant wall and alternately close each eye. The nearer finger shifts more against the background than distant features do. A satellite mapping system applies a calibrated geometric model to related viewing differences. The analogy explains parallax — open definition; it is not a formula for satellite elevation, and a pair of images without known geometry cannot provide an accurate height merely by being called stereo.

Keep the named branches and products distinct

In an invented validation exercise, surveyed elevations at three checkpoints are 100, 150 and 200 metres. A terrain model reports 103, 153 and 203 metres. Each residual, defined here as model minus survey, is positive three metres. The consistency suggests a systematic offset to investigate. It does not establish the cause, or prove that subtracting three metres everywhere will fix areas that were never checked.

Use independent checkpoints rather than only the points that helped fit the model. Also distinguish a surface model that follows roofs and treetops from a terrain model intended to describe the ground beneath them. A forest canopy can create a legitimate difference between two products answering different questions. Ziyuan’s resource, spectral and mapping branches make more sense when their identities, measured quantities and validation methods stay connected.