Key natural impacting factors of China's human population distribution

China is a large country with a population of 1.292 billion, accounting for 115 of the world population. However, the geographical distribution of population is greatly unbalanced. There are many factors contributing to the unbalanced population distribution, including natural condition, social and economic development, and transportation infrastructure. In this article, we specifically explore how natural factors (resource and geographical factors) influence the population distribution. We use the multiple linear regression to analyze how the county population and prefecture (districts under city administration) population density is influenced by different natural factors. The analysis is conducted at two scales: the national scale and the regional scale (including east region, northeast region, middle region and west region). Our results show that arable land proportion, precipitation, elevation, accumulated temperature (also called Effective Accumulated Temperature (EAT), means the sum of daily temperature for the period of daily mean temperature (Td) above 10 degrees C) and the distance from energy distribution location, the distance from main rivers are the significant factors for all the counties in China. For the prefectures, the significant impacting factors include elevation, arable land proportion and distance from the main rivers. For the east region, elevation and arable land proportion are the significant impacting factors of population distribution. For the northeast region, arable land proportion, precipitation and accumulated temperature are the significant impacting factors. For the middle region, arable land proportion, accumulated temperature, distance from energy distribution location and elevation are the significant impacting factors. For the west region, arable land proportion, precipitation, distance from the main rivers, distance from energy distribution location and accumulated temperature are the significant impacting factors.