Use space data before building a satellite

Bangladesh’s space applications have roots in a ground station established in 1968 to receive automatic picture transmissions. After NASA launched the first Earth Resources Technology Satellite in 1972, a Bangladesh ERTS program developed applications of its observations. The Bangladesh Landsat Programme followed, and in 1980 it merged with the Space and Atmospheric Research Centre to form SPARRSO, the Bangladesh Space Research and Remote Sensing Organization.

This history is about receiving, analyzing and applying observations, not about Bangladesh owning the Landsat spacecraft. SPARRSO’s official remit spans natural resources, mapping, weather and related research. A local institution contributes knowledge of the landscape, measurement validation and contact with the people who need the results. Those capabilities can be developed using shared satellite data, alongside later national spacecraft projects.

The two images here show Bangladesh and surrounding areas through Aqua’s MODIS instrument on 2 June and 25 July 2020. NASA combines infrared and visible measurements so water looks dark blue or black, vegetation bright green, and clouds white or cyan. These are NASA-produced examples of a remote-sensing method, not products attributed to SPARRSO. The colors encode selected spectral bands rather than reproduce an astronaut’s view.

Look for the expansion of dark water in the northeastern part of Bangladesh between the images. Compare matching locations and keep the dates and band combination attached to each scene. Bright cloud-covered areas are not confirmed dry land: the surface is obscured there. A before-and-after pair can reveal a changing water pattern, but it cannot by itself tell you the water depth, damage to each building or the moment every location flooded.

Keep missing observations visible

Imagine a simplified map with 100 equal-area cells. Twenty are classified as water and thirty are obscured by cloud, leaving fifty observed non-water cells. Water occupies twenty percent of the whole map as directly identified, or about twenty-nine percent of the seventy visible cells. Neither calculation proves that only twenty percent of the whole area is flooded. The unobserved cells must remain unknown until further evidence arrives.

An analyst should preserve that uncertainty when handing a map to another organization. Report acquisition time, spatial scale, the classification method and missing observations. Independent ground measurements or another suitable sensor can help validate the interpretation. These invented cell counts illustrate the reasoning behind an application, not a measured result from the displayed images. The durable national capability is the ability to turn observations into useful, qualified information.