Gravity model

A family of models estimating interaction between places from their sizes and separation, often adapting Newtonian gravity’s inverse-distance pattern to flows such as migration or trade.

Spatial interaction: The movement or exchange of people, goods, information, or services between locations. The gravity model is a quantitative way to represent interaction between places.
Newton's law of universal gravitation: A physical law stating that masses attract with force proportional to their product and inversely proportional to squared distance. The place-size-over-distance analogy takes its name and basic form from this law.
Retail gravitation: The use of spatial interaction models to estimate where consumers shop. Retail applications use store attractiveness and travel distance to predict customer flows.
Ernst Georg Ravenstein: A nineteenth-century geographer known for empirical studies of migration patterns. His migration laws described regularities that later gravity models sought to quantify.
Spatial autocorrelation: The correlation of a variable with values at geographically nearby locations. Nearby flows and places can violate independence assumptions in gravity-model analysis.
Distance decay: The decline in interaction or influence as distance increases. It captures the model’s central assumption that separation suppresses flows.
Population potential: A measure of population influence at a location, weighted by distance from other populations. It applies gravity-style distance weighting to populations around a place.
Huff model: A probabilistic retail model estimating the likelihood that consumers choose particular shopping centers. It turns gravity-style attractiveness and distance into probabilities of store choice.
William J. Reilly: An American geographer who formulated a law of retail gravitation in the 1930s. His retail model helped establish gravity reasoning in human geography.
Ecological fallacy: The error of inferring individual behavior from aggregate data. Aggregate place-to-place flows do not directly reveal individual choices.

Linked from 10 pages