Improved Heterogeneity in Urban Climate Simulations from HRLDAS

08/02/2013 - 3:00am to 3:30am
ML - Main Seminar Room
Allison Cullen


The High Resolution Land Data Assimilation System (HRLDAS) has been adapted to simulate temperature in urban environments. The model includes many land use types, including three separate urban types: low density, high density and commercial. These land use types are determined based on characteristics of the buildings in each grid cell, including the fraction of land in each grid cell that is urban, vegetated, water, etc. When the data for the urban grid cells are compared to satellite observations, we see a strong warm bias in the model output. We also see that allowing for heterogeneity in the urban fractions assigned to the grid points is an improvement over the traditional approach using a look-up table with fixed urban fractions. We are currently using statistical approaches to examine the causes of these differences. In particular, we utilize a linear model to examine the role of urban fraction and interaction of land use type and the forcing data. Motivated by the role of the simulations in a broader study of the impact of heat stress and the urban heat island on public health, we focus on measurements at both midnight and noon during the summer, defined as June through August.