A Novel Method for Saturation Effect Calibration of DMSP/OLS Stable Light Product Based on GDP Grid Data in China Mainland at City Level

景欣

DOI: https://doi.org/10.3969/j.issn.1672-0504.2017.01.007

2017-01-01

Abstract:DMSP/OLS stable light data reflects various aspects of human activities.It is useful to monitor socioeconomic phenomena and has been widely used in several areas,such as urbanization,air pollution,power consumption surveillance,disaster detection,population estimation,etc.,and the nighttime data can even do better than conventional methods in some specific conditions.However,since the OLS sensor has no onboard calibration system,the DMSP/OLS stable light data have only six-bit quantization and a limited dynamic range of 0 ～ 63.The data are normally saturated over bright urban areas,which limits the usefulness of their application.In China,there are significant saturation effects in many urban areas and the information miss in these saturation areas.Thirty-six saturated cities in China mainland were selected as the study area,and each city had either more than 200 saturated pixels or 1％ higher ratio of saturated pixels.Then a novel approach using Chinese 1 km2 geospatial distributions of Gross Domestic Product (GDP) grid data from 2010 to reduce the saturation effect in China at city level has been presented.The results reveal that the Sum Of Light (SOL) in all selected cities increased reasonably after using the GDP auxiliary calibration method.To validate the new method,two socioeconomic parameters (Built-Up Area (BUA) and Power Consumption (PWC)) were selected for correlation analysis with the corrected light data.The correlation coefficients between corrected SOL and BUA and PWC increased from 0.63 and 0.57 to 0.67 and 0.61,respectively.Thus,the GDP auxiliary method can be applied effectively to conditions in China.