The Influence of Landcover Parameterization Schemes on WRF Simulations of Neighborhood-Scale Heat Stress in Baltimore, Maryland

Buildings, greenspaces, waterways, and other manmade structures within urban areas create a complex landscape and alter the local surface energy balance. The highly heterogeneous urban landscape can cause heat stress to vary within a city, and this is difficult to represent accurately in numerical models. Studies have addressed surface heterogeneity in cities by representing the subgridscale landcover via increasing the model resolution, applying a landcover mosaic scheme, or subdividing the landcover into more categories or local climate zones. Despite the development of these approaches, data-based, multivariate, long-term evaluation of the efficacy of these numerical methods are limited.

In this work, we improve landcover representation within a model and evaluate neighborhood scale heat stress within Baltimore, Maryland during the summer. First, we create a unique location-specific World Urban Database and Access Portal Tools (WUDAPT) L0 landcover dataset that is verified independently with building-level data and place-based experts who understand the morphology of the city. The Weather Research and Forecasting Model (WRF) is then used to create a suite of four model runs that simulate heat stress during the summer of 2022 in Baltimore with 30-meter National Landcover Database (NLCD) data, a custom high-resolution WUDAPT dataset, and landcover datasets with mosaic tiling schemes applied. Finally, a large heterogeneous network of observations, consisting of surface-based data from the NOAA Meteorological Assimilation Data Ingest System (MADIS), MODIS skin temperature measurements, crowd-sourced weather stations, and an array of personal weather stations, is used to evaluate the model.

Overall, the aim of this study is to show the importance of landcover representation in modeling neighborhood-scale heat stress. This work is a prelude to a more comprehensive evaluation of the modeling system that will include examination of the simulated vs. observed surface fluxes and boundary layer development, both of which will also be quantified as part of the Baltimore Social-Environmental Collaborative.