Choose detail, bands and coverage

SSTL described NigeriaSat-2’s fine-imaging mode as 2.5-metre panchromatic and 5-metre multispectral sampling across a 20-kilometre swath. Panchromatic measurements combine a broad wavelength range into one intensity channel; multispectral measurements keep several wavelength ranges separate. The two sampling values describe different products. They do not mean that every independent colour measurement was acquired at 2.5 metres.

For an ideal square grid, a 5-metre sample covers 25 square metres and a 2.5-metre sample covers 6.25 square metres. Halving the spacing produces four times as many samples over the same area. But whether a small object can be identified also depends on contrast, blur, geometry and processing. A feature occupying one sample is not automatically a recognizable building, vehicle or damaged roof.

NigeriaSat-X was designed for 22-metre multispectral sampling across about 600 kilometres. NigeriaSat-2 also had a wider-area 32-metre multispectral imager across about 300 kilometres. These modes address different combinations of coverage and detail. A broad regional survey and a close examination of one urban district need not use the same product.

In a simplified mapping exercise, a 600-kilometre-wide region takes thirty adjacent 20-kilometre strips to span, before allowing for overlap or unusable imagery. A 600-kilometre swath spans that width in one strip, but at a different sampling scale. This arithmetic is a geometric comparison, not a claim that a real satellite can acquire and deliver every required strip in one pass. Clouds, orbit access and tasking still matter.

Halving sample spacing changes both dimensions

A sharp-looking colour display may combine a finer panchromatic channel with coarser multispectral data. Such a product can improve visual presentation, but the added spatial detail is not an independently measured fine-resolution spectrum at every displayed pixel. Record which source channels and processing steps produced the image before using it for quantitative interpretation.

For a fictional 100-metre by 100-metre square, a perfectly aligned 5-metre grid has 20 by 20 samples: 400 in total. A 2.5-metre grid has 40 by 40, or 1,600. The increased sample count is useful information about storage and spatial sampling. It does not by itself prove four times better classification accuracy or four times more reliable damage estimates.