The Correcting Of The DMSP/OLS Datasets And Population Spatialization For China And The USA

Generally, the traditional census data, based on the administrative boundary, cannot reflect the spatial distribution heterogeneity of population. Nevertheless, gridded population spatialization, which provides the more accurate and intuitive population aggregation details in space, can solve the problem and has become an important method in the research of population distribution. However, gridded population spatialization always needs social economic data, which is difficult to acquire, adding the complexity to the study.The Night Time Light (NTL) data, acquired by the Defense Meteorological Satellite Program/ Operational Linescan System(DMSP/OLS), is a key indicator of human activities and has been widely applied to study the population distribution. However, the NTL datasets acquired from different sensors in the same year are incompatible. The NTL datasets acquired in different years are discontinuous. In addition,the NTL datasets has problems such as geographic errors and light overflow. At present, there are no relatively systematic methods provided to save these problems.In this study, the China and the USA were selected as the study area.The inter-calibration, geometric correcting, saturated correcting and light overflow correcting were applied to the NTL datasets in 1992-2013.Based on the NTL data, Land-Use and Land-Cover Change (LUCC) data,county vector data, census in 2000 and 2010, the gridded population spatialization and modification method were executed. In addition, the population distribution of China and the USA were analyzed.The research mainly got the following conclusions: (1) The NTL data correcting methods proposed in this study are available to solve or weaken problems such as data saturation, incompatibility and discontinuity. Using the correcting methods, the more accurate,large-scale, long-time series NTL products can be accomplished. (2) The NTL products are able to be applied to the gridded population spatialization. Additionally, the LUCC data connects the population space distribution and geographical environment factors, improving the precision of the gridded population distribution data. (3) The similarities of population distribution in China and the USA are: ? Population density is lager in the coastal areas than the inland. ? Population distribution is unbalanced between the east and the west. ? In the period of 2000-2010, the total population increased, and part population in the east distracted to the west. The differences of population distribution in China and the USA are:? For the USA, human distracted from the high population density areas to the low periphery.The population density reduced in lots of high density areas whereas the surroundings, especially many small cities, gained a higher population density.? For China, human expanded from the high population density areas to the low periphery. However, the population density of most high density areas had even not reduced.This paper provides effective reference and scientifically basic information for the study of the population distribution, and is helpful for the related department to solve resource utilization, execute disaster evaluation, accomplish environmental management and coordinate the human-earth relationship. The study has great significant on researches of population expansion and distraction, inter-relationship between population and environment, as well as the ecological environmental protection.