Evaluation of interpolation methods for the prediction of urban methane concentrations

Having data about atmospheric concentrations in an entire urban area is difficult, hence inter-polation methods are helpful. Their choice will depend on minimising the error. In this work, two deterministic (Inverse Distance Weight , Local Polynomial Interpolation) and two stochastic methods (Simple and Ordinary Kriging) were applied to predict seasonal and annual atmospheric methane (CH4) concentration means. Two sampling networks were designed in an intermediate city, covering a wide variety of urban densities, with different sampling site numbers. The main objective was to find the interpolation model that best predicts CH4 concentration and to analyse if the network's expansion improves the metric errors -the mean error (ME) and the root-mean -square error (RMSE). The ME values were close to zero in all cases , the stochastic methods had the smallest RMSE for both networks. Besides, adding more sampling sites improved up to 50% of the RMSE values. Finally, an integrated map was obtained incorporating all the best interpolation models, which gave a difference of <4% between the measured and the estimated CH4 concentration. This type of study is helpful to evaluate the design of a sampling network, the territorial planning and future installations of CH4 sources.

Automated summary

Interpolation methods are used to predict atmospheric methane concentration means in an urban area. Two deterministic and two stochastic methods were applied to analyze the metric errors and the impact of adding more sampling sites. The results showed that the stochastic methods had the lowest root-mean-square error and adding more sampling sites improved the accuracy of predictions.