Uncertainty

When we convert information from the real world into a geospatial dataset, we are forced to simplify it, so as not to be overwhelmed by the detail. This simplification might involve:

• Generalising - combining different categories to reduce thematic detail
• Degrading resolution to limit temporal and spatial detail

The act of simplifying therefore introduces uncertainty, which is compounded by our inability to collect accurate measurements of many natural phenomena. This means our data are an imperfect representation of the world, both deliberately and through limitations in our observational capability

Uncertainty may include:

• Thematic uncertainty - where do you draw the line between open forest, and clearing with scattered trees?
• Spatial uncertainty - difficulties defining the location or extent of features due to fuzzy vegetation boundaries, or a rapidly-changing situation on the ground
• Temporal uncertainty - population size, vegetation cover and weather patterns may all change seasonally or through time. Our snapshots of field data don’t necessarily reflect the current situation, and may not coincide temporally with our other datasets

---