Leonelli and Tempini examine how radically heterogeneous health, climate, environmental and biomedical datasets can be linked without pretending that their original contexts are equivalent. Their iconic operation is the strategic invariant: a parameter such as geolocation is temporarily held stable so that otherwise incomparable data can be searched, linked and interpreted together. The theoretical contribution is a situated account of invariance in which comparability is deliberately constructed and its limitations remain visible. Rather than treating localisation as an external universal coordinate, the paper insists on its idiographic character and on documenting the transformations required for reuse. Methodologically, the argument reconstructs big-data practices at the intersection of epidemiology and environmental research, showing how metadata, scale, time and purpose condition inference. The wider bridge is exceptionally relevant to spatial research: built and natural environments can enter health analysis only through carefully governed translations between epistemic cultures, making scale itself an evidential problem rather than a neutral container.
