Abstract
Wind turbine aeroacoustic noise remains an important constraint on onshore turbine siting and operation. The Brooks-Pope-Marcolini (BPM) model provides a computationally efficient semi-empirical description of airfoil self-noise, but its baseline formulation does not explicitly account for atmospheric inflow turbulence. This paper introduces an additive turbulence correction motivated by rapid distortion theory and parameterized by turbulence intensity I_u and integral length scale Λ. The correction is calibrated against turbulent-inflow airfoil data and evaluated using field measurements from a four-turbine onshore wind farm in Kazakhstan. Across 312 ten-minute periods, the standard BPM prediction shows a mean bias of -3.2 dB(A) and an RMSE of 4.1 dB(A), whereas the modified formulation yields a +0.4 dB(A) mean bias and a 2.6 dB(A) RMSE, corresponding to a 37% reduction in RMSE and a 2.8 dB(A) reduction in the magnitude of mean bias. The results indicate that explicit characterization of inflow turbulence can improve low-order wind turbine noise predictions, while the remaining error under stable stratification highlights the need for stability-dependent turbulence representations.