Abstract: Through establishing the STIRPAT model of water footprint changes, the impact of population, GDP per capita, and ten thousand yuan of water footprint changes from 2002 to 2014 was quantitatively analyzed, and the spatial spillover effects of these three types of drivers in 2014 were analyzed by geographically weighted regression. The results showed that from a time point of view, the water footprint and the water footprint intensity showed inversely proportional relationship, that was, the water footprint intensity decreased with increasing water footprint; Population, GDP per capita had the greatest impact on the water footprint changes, which played a positive effect while the ten thousand yuan of water footprint had least impact, but it had an inhibitory effect on the water footprint. From the spatial point of view, each administrative district was affected by the population, GDP per capita and ten thousand yuan of water footprint effects, and the spatial distribution of water footprints showed significant differences.

Key words: water footprint, driving factors, STIRPAT model, geographic weighted regression