This study aims to create a fine grained mapping of the migrant population in Istanbul using land use, nighttime satellite, and extended detail records (xDR) data. We use statistical bias correction methods such as calibration and weighting, spatial scaling methods, and machine learning methods to create the fine granular maps. The use of big data allows for a granular analysis of migrant behavior, contributing to evidence based policies, which can improve the living conditions of migrants. In this study, we use only aggregated data in order to protect personal data. The results demonstrate that satellite and mobile data sources can be used for fine-grained population mapping.