Jinxin Wang-April 28

Urban Heat Island in the Dallas-Fort Worth during Dry and Wet Summers: An Evaluation of Single-layer and Multi-layer Urban Canopy Models

The urban heat island (UHI) intensities in the Dallas Fort-Worth (DFW) region in a dry summer (July 2011) and a wet summer (July 2015) are investigated using observations from Texas Commission on Environmental Quality (TCEQ) stations. It is found that the daytime UHI intensity of wet summer is stronger than that of dry summer, with a positive daytime UHI in wet summer and negative daytime UHI in dry summer. Multi-regression indicates that the urban-rural dew point difference is a significant indicator of day-to-day variation of daytime UHI intensity and the mean wind speed is a significant indicator of day-to-day variation of nighttime UHI intensity. To further understand the reasons for the observed behaviors of UHI, numerical simulations with urban canopy models (UCM) are performed.

Currently, the WRF model has one single-layer UCM (SLUCM) and two multi-layer UCMs. The single-layer UCM has been extensively applied to urban studies and can be coupled with both local planetary boundary layer (PBL) (e.g. MYJ scheme) and non-local PBL schemes (e.g. YSU scheme). The two multi-layer UCMs are the building environmental parameterization (BEP) model and the building energy model (BEM), which are currently coupled with local PBL schemes only. The BEP model does not reproduce the nighttime UHI well and is therefore not used in this study. In this study, three UCM+PBL configurations are tested: SLUCM+YSU, SLUCM+MYJ, BEM+MYJ.

Daily simulations are performed over the two summers, starting from 00 CST NARR analyses one day earlier, with the first 24-hour as spin-up and the late 24-hour for analysis. The results indicate that the PBL scheme has a stronger control on UHI than the UCM scheme. The UCM most prominently affects the surface layer, while the PBL scheme affects both surface layer and PBL. For both urban and rural areas, all simulations with the YSU scheme produce cooler surface at daytime and warmer surface at nighttime. The nighttime UHI is more strongly controlled by the PBL scheme. SLUCM+YSU produces weaker UHI than SLUCM+MYJ and BEM+MYJ. For the urban area, BEM+MYJ tends to perform better than SLUCM+MYJ. BEM+MYJ can reproduce the negative daytime UHI in dry summer, while UCM+MYJ produces erroneous positive daytime UHI in the dry summer. Therefore, the daytime UHI simulated by BEM+MYJ is better correlated with the urban-rural dew point difference as indicated by observations. BEM+MYJ produces weaker urban winds, while SLUCM+MYJ erroneously creates stronger urban winds. The weaker near-surface eddy diffusivity simulated by BEM is suggested to be the reason for the weaker urban winds, because BEM produces a more stable surface layer. In both SLUCM and BEM, near-surface inversion is a good indicator of day-to-day variation of nighttime UHI.