Abstract
(a) When environmental variables are included in hedonic house price models, the locations where a property transaction has been taken place are more than those equipped with an environmental monitoring station. (b) When environmental variables are numerous, an environmental quality index (EQI) is needed. The solution is to interpolate (kriging) environmental variables and, subsequently, elaborate an EQI, because of the lower prediction variance. (c) Environmental informations can be both objective and subjective. The potential mismatch between the spatial support for both the objective informations and the subjective ones is solved by using a kriging strategy (area-to-point kriging) that forces scarce objective environmental informations to be coherent with dense subjective ones.
These options for elaborating EQIs are compared in Madrid City (Spain).
