Machine Learning Emulation of Urban Land Surface Processes
This provides a novel approach to improve urban land surface modeling for climate and weather prediction applications, though it is currently constrained by training data from a single site.
The authors tackled the problem of modeling urban land surface processes by developing an urban neural network (UNN) that emulates the mean output of 22 existing models, resulting in greater accuracy relative to flux observations, less computational cost, and fewer input parameters compared to a reference model, and stable, more accurate performance when coupled to a weather forecasting model.
Can we improve the modeling of urban land surface processes with machine learning (ML)? A prior comparison of urban land surface models (ULSMs) found that no single model is 'best' at predicting all common surface fluxes. Here, we develop an urban neural network (UNN) trained on the mean predicted fluxes from 22 ULSMs at one site. The UNN emulates the mean output of ULSMs accurately. When compared to a reference ULSM (Town Energy Balance; TEB), the UNN has greater accuracy relative to flux observations, less computational cost, and requires fewer input parameters. When coupled to the Weather Research Forecasting (WRF) model using TensorFlow bindings, WRF-UNN is stable and more accurate than the reference WRF-TEB. Although the application is currently constrained by the training data (1 site), we show a novel approach to improve the modeling of surface fluxes by combining the strengths of several ULSMs into one using ML.