Controls of Hyperlocal Heat Hazard in Miami-Dade County

Heat is an increasingly important issue as our climate changes, particularly in urban areas due to the urban heat island effect. While heat warnings are currently issued by the National Weather Service, they often cover broad regions that fail to capture hyper-local variations in heat hazards.This project focuses on analyzing these localized heat hazards to improve upon the existing warning systems. For example, while conditions at the NWS weather station might not require a heat warning, other parts of the city could be experiencing dangerous heat conditions. By studying the differences in heat indices within a city, we aim to determine whether these measurements are heterogeneous across different locations. One hypothesis is that locations closer to the ocean may have higher humidity values which can elevate heat indices compared to inland locations. This project investigates how humidity changes with distance from the ocean and whether there is a systematic relationship between location and humidity. Additionally, this project will analyze whether the inland temperature increase is enough to offset the humidity difference. If not, humidity could be a primary driver of heat hazard differences. The goal is to develop an algorithm capable of predicting the heat index at various stations around Miami, based on variables measured at the MIA weather station (temperature, dew point, humidity, wind, etc.). This would allow for a more accurate representation of hyper-local heat hazards. Data for this analysis comes from multiple sensors measuring humidity and temperature, as well as WeatherSTEM stations that track a full range of meteorological variables. Combined with data from the weather station at Miami International Airport, machine learning techniques such as regression will be used to aid in these predictions, ultimately enhancing our understanding of localized heat hazards.