Modeling of Noise Pollution in Residential Area and its Health Hazard using Probability Distribution Models

Noise pollution is an excessive sound produced intentionally or accidently which can have deleterious health effects and can reduce environmental quality. This work proposed the modeling of noise level (NL) in a residential area containing block, milling and timber industries. Gamma distribution model (GDM), Lognormal distribution model (LDM) and Weibull distribution model (WDM) were applied to model the NL measured and estimate their respective maximum likelihoods. Goodness of fit (GoF) was determined to select the best fit model. Probability of exceeding critical point (PECP) of the NL at each industry was obtained and the NL estimated was compared with WHO standards. The means of the NL for the three industries were obtained using GDM, LDM and WDM as 50.037dBA, 51.014dBA and 76.747dBA, 51.014dBA, 59.892dBA and 76.745dBA, 58.582dBA, 71.376dBA and 70.886dBA respectively, their corresponding maximum likelihoods (ML) were also estimated as 0.1593, 0.2103, 0.1589, for GDM, 0.1581, 0.1892, 0.1553 for LDM, and 0.1598, 0.2111, 0.1683 for WDM. The PECP were determined as 0.3146, 0.1158 and 0.5471 for block, milling and timber industries respectively. It was found that ‘LDM’ is the best fit model because it has the least GoF and workers from the block industry are at risk of developing mild annoyance, while workers from the milling industry may suffer severe annoyance, mild sleep disturbance and poor next day mode, whereas workers from the timber industry are likely to develop mild hypertension and cardiovascular problem, severe annoyance, poor next day mode, sleeping disorders and poor performance.