Turn an acquisition into a map

SANSA reports SumbandilaSat imagery used for Namibian flood monitoring, Kruger fire work and observations following disasters in Japan and the United States. Those examples motivate a crucial question: what decision could each observation support? A scene can show surface conditions within its coverage at acquisition time. It cannot by itself establish the condition of hidden areas or the state of a fast-changing event several days later.

Consider a fictional river scene observed Monday morning and published Tuesday afternoon. A flood boundary derived from it describes the Monday observation, with the uncertainties of its classification. The publication time tells you when users could obtain the product. Keep both times visible. If responders need Tuesday’s conditions, the Monday image is context that should be combined with newer observations or ground reports.

Image acquisition, onboard storage and ground transmission compete for finite resources. The manufacturer describes primary storage and additional redundant capacity; backup capacity should not automatically be treated as extra space for unrelated observations. Likewise, a headline radio rate is not a promise that every second of a pass delivers useful image bytes. Contact setup, protocol overhead and interruptions affect delivery.

Use a fictional payload with 360 megabytes waiting and a useful transfer rate of 2 megabytes per second. It needs 180 seconds, or three minutes, before any additional tasks. If only two minutes remain, 240 megabytes can arrive and 120 remain. An operator might prioritize a smaller region or use another pass. This is a decimal-unit teaching budget, not SumbandilaSat’s measured link performance.

A receiving window limits delivery

An analyst can compare before-and-after scenes, align them to a common reference, mask cloud and shadow, and examine candidate changes. Ground observations can help test whether a dark patch represents floodwater rather than a shadow or another surface. A cloud-covered pixel should remain unknown unless other evidence resolves it; classifying it as dry creates false certainty.

Before sharing a map, record the acquisition time, processing version, coordinate reference system, valid coverage and method used to classify the surface. Check the actual product metadata when a catalogue summary is ambiguous. A map that separates observed flooding, known dry ground and unobserved areas helps a user choose the next measurement. An attractive image with those distinctions removed can lead to a poorer decision.