Integrating atmospheric fluctuations into the image source method

Successive room impulse response (RIR) measurements exhibit subtle variations, mainly caused by atmospheric fluctuations due to spatiotemporal air-temperature differences. This work extends the image source method by incorporating temperature-induced variability through global deterministic drift and stochastic local volatility. Measurements in a shoebox room show that the proposed model reproduces key behaviors observed in practice, including the structure of differential RIRs, temporal variance of early reflections and coherence of late reverb, which fluctuation-free simulations fail to capture. The method increases realism without increasing the algorithmic complexity, enabling simulations that better reflect the variability encountered by algorithms in real measurement scenarios.