Quantifying Disease Dynamics in Post-Disaster and Acute Epidemic Settings Using Passively Collected Data

Abstract

In Chapter 1 we evaluate the spatial and temporal spread of COVID-19 with varying lockdown parameters informed by mobility patterns derived from mobile phone data. Early in the SARS-CoV-2 pandemic, mobility data sources signaled increased movement immediately preceding a lockdown followed by depopulation of urban centers; both of which can counter the purpose of a lockdown. We developed a spatial metapopulation model, informed by the mobility data, to evaluate the effects of varying lockdowns on the epidemic spread of SARS-CoV-2. We further incorporate real mobility data from Spain to identify potential hotspots of transmission that were potentially seeded from the capital city of Madrid. In Chapter 2 we investigate the transmission dynamics of COVID-19 in the United States using Ad-tech based GPS traces. Mobility data have been used exhaustively in the United States and internationally as a proxy for contact rates which drive the effective reproductive number of SARS-CoV-2. In this study we use GPS data from ad-tech sources to understand the mechanistic link between mobility and the effective reproductive number. We evaluate the reliability of these metrics are proxies for the contact rate and determine spatial and temporal boundaries under which they perform well. In Chapter 3 we augment the public health response to Leptospirosis outbreaks following a flooding event in Kerala, India using remote sensing and mobility data. In Kerala, a southern state in India, an acute flooding event in August of 2018 resulted in a large outbreak of Leptospirosis, which is endemic in the region21. Here we use mobility data and time-varying, state-wide flooding probability maps derived from remote sensing imagery to augment the existing methods and characterize the dynamics of the different Leptospirosis outbreaks in the northern and southern districts of the state.