With the formulation of different policies

special economic zones dramatically expanded in the past 40 years. Basing from remote sensing and geographic information system technology

we performed long-term and high-frequency monitoring of urban expansion in special economic zones by using multi-source remote sensing images between 1973 and 2013. Shenzhen

Xiamen

and Haikou were selected as research subjects. All spatial expansion information was obtained through human-computer interactional digital interpretation. Basing from established-interpretation symbols of urban lands

researchers with experience in visual interpretation referred to Google Earth and topographic maps to ensure the accuracy of monitoring results above 90%. This study selected expansion speed

influences on land use

compact ratio

and centroid shift as indicators

and combined the effects of natural and man-made elements to analyze the similarities and differences of spatiotemporal characteristics among Shenzhen

Xiamen

and Haikou. Four major results were obtained. First

the expansion speed of special economic zones in the past 40 years experienced one low-speed stable stage

two acceleration stages

and two deceleration stages. The expansion speed of Shenzhen was the fastest

followed by Xiamen and then Haikou. This situation fully reflects the significance of national and local policies

as well as social and economic development. Second

1370. 61 km2 nonurban lands around Shenzhen

Xiamen

and Haikou were converted to urban land between 1973 and 2013. Arable land was the first land source of special economic zone urban expansion. Other main land sources of urban expansion include forest land

water body

rural settlement

industrial and traffic lands

and sea area. Grassland and unused land had minimal contribution to urban expansion. Rural settlement and industrial-traffic land was the second land source of Xiamen and Haikou urban expansions

and the third land source of Shenzhen urban expansion. Forest land was the second land source of Shenzhen urban expansion but produced a contribution rate of < 10% to Xiamen and Haikou. Third

the compact ratio of special economic zones decreased. Before 2004

land resources around special economic zones were relatively adequate

urban expansion was fast

land-use efficiency was low

and compact ratio considerably reduced. After 2004

urban expansion space was limited

expansion speed slowed down

and compact ratio stabilized. Fourth

marine reclamation engineering appeared during urban expansions in Shenzhen

Xiamen

and Haikou. Under the combined effects of the policies and marine reclamation engineering

the centroid of special economic zones tended to migrate toward the coastline. Affected by the comprehensive influence of expansion area and compact ratio

Shenzhen and Haikou had the large stand smallest centroid shift distances

respectively. Remote sensing monitoring of expansions in special economic zones can provide support for the future projection and policy formulation of special economic zones.