Satellite brochures love the word resolution, but beginners often treat it as a simple quality score, where higher is always better. In practice, resolution is a trade-off between detail, coverage, cost, and file size, and the right choice depends on the smallest thing you need to see. A 30-metre image can map a region brilliantly and fail completely at a single building. A sub-metre image can show that building and bankrupt a large-area project. Understanding spatial, spectral, and temporal resolution together helps you pick the level your question actually requires.

Spatial resolution is the ground size of a single pixel. Landsat's 30 metres means each pixel averages everything in a 30-by-30-metre square, so a narrow road, a small pond, or a single house blends into its surroundings. Sentinel-2's 10 metres resolves fields, larger buildings, and river bends more cleanly. Commercial sub-metre imagery, often below one metre, can show cars, individual trees, and roof details. As a rule of thumb, a feature should be several pixels across to be recognised reliably, not just one pixel wide. If you need to identify a feature confidently, aim for imagery where it spans at least three pixels, and test with a known example before committing to a large download.

Spectral resolution is how many bands a sensor records and how narrow they are. More bands, especially red-edge and shortwave-infrared, let you separate vegetation types, water, and built surfaces that look alike in ordinary colour. Temporal resolution is how often the satellite revisits the same place, which decides whether you can catch a flood peak, a harvest, or a seasonal change. A sharp image that arrives once a year is useless for weekly crop monitoring, while a coarse image that arrives every few days may be perfect for it. When comparing products, list all three resolutions side by side, because a product that wins on one dimension often loses on another.

Free programmes such as Landsat and Sentinel-2 cover the globe repeatedly at no cost, which is why they dominate teaching and research. Sub-metre commercial imagery is typically licensed per area and per use, so covering a whole province can cost far more than covering a single site, and the licence may restrict redistribution. File size grows quickly with resolution: halving the pixel size quadruples the number of pixels for the same area, which slows downloads, processing, and display. For large areas, start coarse, find where detail matters, and buy or request fine imagery only for those hotspots. That staged approach keeps projects affordable and computers responsive.

Map a region, track decades of change, or monitor a season: start with 30-metre Landsat or 10-metre Sentinel-2 and check whether the patterns you need are visible. Map a neighbourhood, a farm, or a construction site in detail: look for sub-metre or drone imagery for that small area, and use satellite data for context. Always state the resolution in your map notes, because a reader cannot judge a claim about a small feature without knowing the pixel size behind it. Resolution is not a badge of quality; it is a measurement choice, and the best choice is the coarsest level that still answers your question clearly.

Do not chase the sharpest image by default. Match pixel size to the smallest feature you must recognise, weigh spectral and revisit needs alongside detail, and reserve expensive fine imagery for the places where it changes a decision. That discipline keeps satellite projects fast, affordable, and honest, and it leaves you with maps whose claims match the evidence behind them.