Assessment and Removal of Building-Induced Errors from Rooftop Anemometer Observations
RAL scientists are working to develop the knowledge and methods necessary to remove building-induced biases in the wind field measured by anemometers mounted on masts above rooftops. Building rooftops are often selected for siting meteorological sensors for measuring winds in urban environments. These measurements are critical for defining city scales and neighborhood scales in the wind field, and must be sufficiently accurate for assimilation into mesoscale models, for direct input to transport and diffusion (T&D) models, and for use in defining ambient winds for building-aware computational fluid dynamics (CFD) models. However, measurements of wind speed and direction over rooftops are highly sensitive to distortion of the mean flow by buildings. A schematic of the possible distortions of the flow over and around even a simple building shape is shown in Fig. 1. The distortions can cause rooftop anemometer measurements to be biased so that the measurements do not correctly represent the mean flow in the urban environment where the measurement is taken. The immediate consequences of not correcting this error are 1) mesoscale meteorological models may have potentially large analysis and forecast errors; 2) T&D forecasts of plumes from releases of hazardous material will be altered by using biased wind fields; and 3) CFD-models that use these winds as input will produce erroneous results.
FY06 Accomplishments:
The building-resolving EuLag LES model was used to document the effects of buildings on the three-dimensional turbulent structure of the wind field above the rooftops. EuLag model simulations account for a different ambient (upstream) winds (speed, vertical shear, direction relative to building), thermal stabilities, and single buildings with simple geometries (height, shape, orientation relative to the wind). The practical impact of the errors was quantified by assessing the sensitivity of mesoscale model forecasts to the assimilation of corrected, vs. uncorrected winds. These sensitivity experiments were conducted for a Washington, D.C. area, using the building-resolving EuLag LES model, and the NCAR's Real-Time Four-Dimensional Data Assimilation (RTFDDA) mesoscale model. SCIPUFF's T&D model was used to evaluate the sensitivity of T&D calculations to the use of corrected roof-top winds. The wind data can be assimilated by a mesoscale modeling system, and those model analyses and forecasts can be used to drive T&D model. Also, the building wind can be directly used to define ambient conditions for a CFD model, which computes street-canyon winds and T&D. The sensitivity of T&D calculations to root-top wind biases was evaluated using both of these approaches. And finally, sensitivity studies were conducted to determine the optimal anemometer placement over specific buildings to reduce the bias error.FY07 Plans:
With additional funding, the current pilot study will be expanded to include extending the validation of the EuLag CFD model building simulations and use the results of these and other simulations to create a catalog of rooftop corrections for various flow structures and building shapes. Additional sensitivity studies for selected cases would also be conducted and the results of the studies used to include more building shapes and wake effects from nearby buildings. These results would be used to create a prototype catalog that would be general enough to use over the entire Washington DC area. The operational bias correction system for DCnet would continue to be evaluated and refined by performing statistical studies of the effect of the bias corrections on mesoscale and T&D modeling in more general urban settings. This will involve further sensitivity studies of both the RTFDDA assimilation strategy and the wind bias correction scheme. The resulting catalog should be general enough to be used to provide recommendations for rooftop placement of new sensors, or relocation of existing sensors in urban areas.