A Comprehensive Assessment of Land Surface - Atmosphere Interactions in a WRF/Urban Modeling System for Indianapolis, IN

The accuracy and biases of mesoscale atmospheric models are important to quantify given the importance of these models for atmospheric transport applications. As part of the Indianapolis Flux (INFLUX) experiment, the accuracy of the Weather Research and Forecasting (WRF) model was assessed. The main goal of INFLUX is to estimate greenhouse gas (GHG) emissions from Indianapolis at high resolution using atmospheric methods. These methods depend on high quality atmospheric transport fields and accurate quantification of atmospheric transport errors. INFLUX provides an opportunity for extensive observation-to-model comparison in order to assess model errors for the following meteorological variables: latent heat and sensible heat fluxes, air temperature near the surface and in the atmospheric boundary layer (ABL), wind speed and direction, and ABL depth. The model-data comparisons will enable the development of modeling systems that more accurately capture the meteorological complexity of urban environments.

The sensitivity of the WRF modeling system to different model realizations was tested by creating an ensemble of runs utilizing different ABL schemes, urban canopy models (UCMs), and a new subroutine that was created to reduce the overestimation of urban land cover. Urban land cover is just one example of the input data used by UCMs in WRF. These input data can be improved by integrating new and updated databases describing urban areas. Updates to the urban land cover parameter improved the modeled sensible and latent heat flux by as much as 30%; however, the bias and spread of the errors in other meteorological variables, such as ABL height, wind speed, and wind direction, were dominated by the parameterization schemes chosen for each model run. While a more accurate representation of urban parameters will not always produce smaller forecast errors, adding more accurate and heterogeneous parameter values will allow us to identify what parameterizations are deficient in the UCM and in the WRF modeling system. By using the observations made available through the INFLUX project, the assessment of model transport errors will lead to a more accurate atmospheric inversion that will be produced by the INFLUX project and a greater understanding of sources of modeled atmospheric transport errors over urban environments.