Dynamic Estimates of Population via the Cadastral Expert Dasymetric System

Keywords: population, group quarters, California, CEDS, housing units, estimates
Abstract Type: Paper Abstract

Abstract

The quality of top-down gridded population models depends on detailed administrative boundaries and high-resolution supporting data. Because highly detailed ancillary data is not consistently available at geographies of scale, gridded population models rarely exceed 100m resolution. However, the increasing availability of remotely sensed data, high-quality group quarters and cadastral information make sub-census estimates more feasible. This research proposes the use of an adapted Cadastral Expert Dasymetric System (CEDS) to validate and select from a suite of distributive models. The result is a patchwork of dasymetric methods for the State of California, that is sensitive to the quality of inputs at local scale. The work is bolstered by the incorporation of a random forest model of housing density, which was found invaluable when capturing recent construction and rural areas where metadata is lacking. Model validation is explored across three California case studies and normalized to statewide coverage. Estimates are adjusted to E1 City/County estimates by the California Department of Finance Demographic Research Unit and made available at increasing scale via the US National Grid.