Prediction of Ambient PM10 Concentration in Malaysian Cities Using Geostatistical Analyses

Abstract

Unplanned urbanisation and industrialisation caused deterioration of the atmosphere, hence affecting human health and the environment. Malaysia is one of the Southeast Asian countries that struggle significantly with air pollution. Although some measures have been taken to monitor the air pollutants level in various cities, the limited monitoring stations in the country leave some areas unmonitored. A spatial interpolation technique is a primary method for monitoring unmanned regions to develop a better mitigation strategy towards reducing urban air pollution. Particulate Matter 10 (PM₁₀), one of Malaysia's primary air pollutants, is used to predict and indicate the air pollutant’s presence in some unmeasured locations. Spatial interpolation models and geostatistical analysis such as ordinary kriging (OK), universal kriging (UK), and Inverse Distance Weighting (IDW) were used in this study to predict and assess the distribution of PM₁₀ to other regions. The PM₁₀ thematic map produced by IDW has a maximum and minimum value of 76 μg/m³ and 42 μg/m³, while the UK predicted a maximum and minimum value of 61 μg/m³ and 54 μg/m³, respectively. The OK predicted a maximum value of 60 μg/m³ and a minimum value of 53 μg/m³. The predicted values from the three interpolation methods aligned with the Malaysian air pollution index (API) with good and moderate air pollution levels. Comparatively, the performance of IDW is more reliable considering its non-bias in predicting PM₁₀ concentration. Conclusively, the air pollution prediction map from this study could be leveraged to control and monitor air pollution, especially in unmanned areas.