Understanding land use evolution for vacant residential and vacant commercial parcels in Miami, Florida

In this study, we develop quantitative models to identify land use changes to individual parcels in Miami, FL. We focus on indicators for changes in population and economic activity. In our analysis, we identify the factors that promote this activity by examining how residential and commercial parcels are changing over time. While significant efforts have been conducted in earlier literature, there is limited work on understanding how parcels of a particular land use type change over time. The main aim of the proposed research effort is to contribute to land use evolution models that are customized by land use type. These frameworks can allow for quantitative exploration mechanisms for policy makers to proactively encourage sustainable adaptation due to emerging challenges associated with global climate change. The frameworks were developed using high-resolution parcel level data from the Florida Department of Revenue for the time period 2011–2020. The framework has two components: (1) a binary logit model to analyze what factors affect a change from the current land use designation and (2) for parcels selected for a change, a multinomial logit model, to examine the factors contributing to the new land use type. We built a joint model system that allows for a simultaneous estimation of the two decision processes. The joint model was estimated using a comprehensive set of independent variables including socio-demographic variables, economic development variables, and transportation infrastructure variables. The models developed were validated using hold-out sample data to show the value of the proposed model system