Build and deliver an image

SUNSAT’s high-resolution imager used three linear CCD channels. A line sensor samples across the view; spacecraft motion supplies successive lines along the ground track. Pointing changes can shift or distort their alignment. The preflight imaging paper therefore treats attitude knowledge, stability, optics and data handling as one connected measurement system. A sharp laboratory image alone does not prove that the complete orbiting chain will work.

The instrument’s green, red and near-infrared channels sampled different wavelength ranges. Near-infrared is outside human vision, so a displayed three-channel picture need not reproduce visible colours. The university’s returned-image gallery describes 15-metre sampling and publishes reduced sections of a larger swath. A screen preview is therefore not the complete original measurement, and the display size is not a new sensor resolution.

For an invented teaching image, use 1,000 samples per line, three channels and one byte per sample in each channel. One line contains 3,000 bytes. Two hundred lines contain 600,000 bytes before metadata or compression. At an assumed useful delivery rate of 1,000 bytes per second, that image needs 600 seconds of contact. These simplified decimal units are not SUNSAT’s operating specifications.

If a usable pass lasts only five minutes, its 300 seconds carry 300,000 bytes at that same assumed rate. The operator could send a reduced preview, choose a smaller region, compress the data appropriately or use another pass. Choosing a region preserves detail there while sacrificing coverage. A preview preserves broad context while reducing detail. Those choices answer different scientific questions.

The downlink limits the image you can deliver

The 1 July 1999 status report says the high-speed image downlink had failed after three uses. The team used a coarse image to select smaller full-resolution sections for slower download. That account matters more for this operating period than a preflight headline transmission rate. The TV path still worked, so one S-band failure should not be described as the loss of every S-band function.

Suppose a flood analyst chooses a small region around a bridge. Its detailed image may support a local comparison, but it does not establish the condition of an entire river basin. Preserve the selection rule and the wider missing coverage. SUNSAT’s example shows how instrument capability and communication capacity jointly determine the evidence that reaches the ground.