Settlement morphology transition is constrained by population migration, settlement satisfaction, and low-carbon city policies. Many land use models poorly describe these transitions, rendering studies in this field relatively inadequate. Xiamen Island, a rapidly urbanizing area in the southeast of Fujian Province in China, was selected as a case study for settlement morphology transition. We generated a system dynamics (SD) model of the carbon footprint of this settlement based on settlement morphology transition and simulated the transition of settlement morphology using this SD model. In this model, the changing social needs of a residential area over the next 20 years were simulated, and the settlement carbon footprint was calculated under three different scenarios (basic, low-carbon, and compact city). We used the results of these simulations as the input of the non-spatial demand module in the CLUE-S (Conversion of Land use and its Effects at Small extent) model. In the study of settlement spatial allocation using the CLUE-S model, we simulated the transition of settlement morphology under three scenarios and depicted the spatial conversion of land use on Xiamen Island over the next 10 years. Next, we discussed the allocation characteristics of settlement types II and type III, which are predicted to be the major settlement types in decades. In summary, this study attempted to depict the transition of settlement morphology on Xiamen Island in multiple dimensions-time, space, quantity, and sequence. To address these issues, we clarified the development characteristics of the settlement in different ways, such as driving forces, structures, spatial allocation, and using various scenarios. The simulation results showed that the SD and CLUE-S models bridge the gap between non-spatial land use policies and the spatial characteristics. The simulation results of the SD model revealed that 1) the ratio of three types of settlement, I, II, and III, was 1:1.18:0.83 in 2009; type III will become predominant in 2018. 2) Over the next 20 years, the total population and the demand for land resources on Xiamen Island will continue to increase. The carbon footprint of Xiamen Island under different scenarios varied significantly. 3) In the basic scenario, the total population of Xiamen Island was predicted to reach 2.40 million in 2030, while the total settlement area was predicted to be 75.3 km². The carbon footprint was predicted to increase and reach a plateau phase, at 4605 kg C/cap, in 2022. 4) In the low-carbon scenario, the total population of Xiamen Island was predicted to reach 2.44 million in 2030, while the total settlement area was predicted to be only 66.4 km². The carbon footprint was predicted to increase and reach a plateau phase, at 3771 kg C/cap, in 2016. 5) In the compact-city scenario, the total population of Xiamen Island was predicted to reach 2.52 million in 2030, while the total settlement area was predicted to be 84.7 km². The carbon footprint was predicted to increase and reach a plateau phase, at 5759 kg C/cap, in 2028. The simulation results of CLUE-S model indicated that settlement type II tends to distribute in areas with high population density, convenient traffic, and proximity to the city center, which can be summarized as the inertial development mode. In contrast, settlement type III tends to distribute in areas with low population density, low elevation, and proximity to commercial centers, which can be summarized as the low-land-cost development mode. Finally, we provide some recommendations to facilitate a balance between the low-carbon and compact-city scenarios. We hope that this study will contribute to the formulation of energy-saving measures as well as the construction of new settlement types. We also hope that it will provide scientific support for the sustainable development of urban settlements.