Over the past 50 years, dengue has been expanding globally into previously unaffected areas. This has been attributed to climate change, urbanisation, and increased connectivity driven by human movement. In Brazil, the rapid expansion of dengue has led to outbreaks occurring in previously unaffected areas, including the temperate South region and remote areas of the Amazon rainforest. In this thesis, I use spatial models to explore the drivers of dengue re-emergence and expansion in 21st century Brazil. Spatial modelling techniques are used to disentangle the effect of increasing temperatures in South Brazil and the contribution of human movement around the Brazilian urban network to the expansion of the dengue transmission zone. First, using Bayesian spatiotemporal models, I found an increased odds of dengue outbreaks in highly urbanised, well-connected cities which had already experienced an outbreak and had year-round temperatures suitable for dengue transmission (Chapter 2). Although these models were able to capture the significant impact of temperature on the expansion of dengue into South Brazil, they were unable to quantify the role of human movement in dengue expansion. I conducted a systematic review to identify how spatial connectivity had been accounted for in models of mosquito-borne disease transmission and the assumptions made about how spatial connectivity arises (e.g., human movement between regions) (Chapter 3). Although the number of spatial modelling papers had increased rapidly over the past 5 years, very few statistical models considered connectivity arising due to human movement and there were no models identified capable of accounting for multiple sources of spatial connectivity. Expanding current state-of-the-art statistical frameworks using ideas from network-based mechanistic models identified in this systematic review became the focus of the remainder of the thesis. I developed a novel statistical modelling approach which can include multiple sources of spatial connectivity, such as similarities between close areas and human movement, and quantify the relative contribution of each source to the overall spatial structure of the model outcome (Chapter 4). This framework was applied to dengue outbreak data for the whole of Brazil between 2001 – 2020 (Chapter 5). Model results showed that human movement based on commuting for work or education contributed very little to the overall spatial structure of the number of outbreaks per municipality in Brazil, but this contribution was significantly higher in North and Northeast Brazil compared to South Brazil. In this thesis, I have explored the complex, interacting drivers of dengue expansion in Brazil since 2001, including increasing temperatures in South Brazil and connections between cities arising from human movement around the Brazilian urban network. Although this thesis focuses on dengue expansion in Brazil, the methods presented here are flexible enough to be applied in any Bayesian hierarchical model where spatial connectivity exists within the data. Given the increasing risk of future pandemic pathogens due to increasing climate and globalisation, robust modelling tools are essential to gain better understanding of infectious disease emergence and identify areas at future risk of expansion.