Understanding the drivers of viral outbreaks is crucial for assessing future epidemic potential and options for outbreak control. However, disentangling the role of intrinsic drivers, such as population immunity, from extrinsic drivers, such as climatic variation or human behaviour, is challenging, particularly in complex settings with heterogenous immune landscapes. In this thesis, I integrate serological and climate data streams within modelling frameworks to disentangle the impact of immune, climatic and behavioural drivers of acute viral outbreaks, and consider the implications for disease control. I focus on SARS-CoV-2 and dengue virus as case studies, two pressing health threats with complex epidemic dynamics and the ability to cause large outbreaks with the potential to overwhelm health systems. Through analysis of a seroepidemiological workplace cohort in the United States from April 2020 – February 2021, I found that primary infection with SARS-CoV-2 provided protection against reinfection for most individuals. I then used a multi-strain, age-stratified compartmental model to estimate the impact of immunity, population behaviour and vaccination on SARS-CoV-2 transmission in the Dominican Republic from 2020 – 2022. By jointly fitting to serological and surveillance data, I found that epidemic dynamics were largely driven by the accumulation of post-infection immunity but that, despite this, vaccination was essential in enabling a return to pre-pandemic behaviour without incurring considerable additional morbidity and mortality. Next, I used a Bayesian hierarchical mixed model to quantify the effects of immunity and climate on dengue dynamics in Singapore from 2000 – 2023, with a view to improving early warning for outbreaks. I found non-linear and delayed impacts of climatic variation on dengue risk, with increased risk at intermediate temperature and rainfall levels, during El Niño events, and in the period following a switch in dominant serotype. I then adapted this model into an early warning framework, generating dengue forecasts at 2-8 week lead times. Accounting for serotype dynamics as a proxy for immunity, as well as climatic variation, improved the predictive power of the forecasting model, and particularly the prediction of outbreaks. Finally, I used a Bayesian space-time hierarchical model, fit to surveillance data from 2013 – 2023 in the 155 municipalities of the Dominican Republic, to estimate the role of climatic and epidemic drivers in spatiotemporal dengue dynamics. I leveraged serological data to estimate a proxy for the build-up of immunity in a dengue season and then accounted for this, as well as autocorrelation in case counts from the force of infection, within the modelling framework. I found evidence of increased dengue risk at higher maximum temperatures and humidity, as well as in drought or El Niño conditions. I found that El Niño and drought indicators are influential predictors of temporal dengue dynamics, while lagged cases, weighted to represent the force of infection, predict both spatial and temporal patterns. Overall, in this thesis, I consider immune drivers alongside important extrinsic drivers within modelling frameworks to better characterize acute viral transmission dynamics in complex epidemiological settings. I found that, by including multiple data streams within modelling frameworks, I was able to distinguish between epidemiological hypotheses underlying disease transmission and improve prediction of outbreak risk to inform epidemic response.