The Internet of Things (IOT) is a concept which emerged simultaneously with developments of data acquisition and sharing technologies, and growth of Information and Communication Technology (ICT). It also quickly involved different aspects of human being life in the modern world. In this study, the problem was investigated of environment monitoring issue, in particular air pollution.

 Air pollution is one of the main problems of the big cities and its harmful effects on environment and public health is inevitable. In this regard, real-time and optimal monitoring of air pollutants concentrations plays a significant role in inhibition of this problem. In this study, by integration of GIS and Internet of Things (IoT), guidelines were presented for optimal management of air pollutions

High temporal and spatial variability of pollutants implies importance of using in-situ sensors to monitor this phenomena. Deployment and maintenance of air pollution sensor networks are very expensive. Moreover, measurement of air pollution concentration at anyplace is also impossible practically.  Considering air pollution monitoring as a scenario in IoT world, accessing concentration level of air pollution at right time and right place, and due to client request is necessary. Therefore, a real-time monitoring system using geoprocessing services is needed which estimates concentration level of pollutants throughout the city. Currently, air pollution monitoring network of Tehran is governed by two organizations, while better estimation of air pollution requires both network data. Another problem is the heterogeneity of this network which make difficult to expand this measurements into another scenarios and projects, and transform information between IoT components in IoT context. Providing interoperability in this network, a spatial data infrastructure (SDI) is required to close these gaps. For this purpose, OGC Sensor Web Enablement (SWE) standards including Sensor Observation Service (SOS), Sensor Model Language (SensorML) and Observation and Measurement (O&M) were used. In proposed monitoring system, RESTful web services in Service Oriented Architecture (SOA) were implemented. Providing reliable information for users in the shortest possible time is another challenge in the Internet of Things. Accordingly, reliability and time as two indicators for evaluation quality of geospatial web services were proposed. Thus, quality of four interpolation services were investigated by utilizing two deterministic methods (Inverse Distance Weighted (IDW) and Global Polynomial Interpolation (GPI)), and two geostatistical methods (Empirical Bayesian Kriging (EBK) and Ordinary Kriging (OK)).

In this study, quality behavior of interpolation web services was examined via study of the six parameters impacts on qualitative indices. Results of this analysis were independent of the measurements data type and could be used in other sensor networks. Among these six parameters, four of them were related to spatial distribution and network structure including average, minimum and maximum distance between in-situ sensors and also number of stations. 2 other parameters were average and standard deviation of sensor measurements data that indicate statistical characteristics of sensor measurements data. Impact of mentioned parameters on indices were investigated by Pearson correlation coefficient.

 Results showed that average RMSE of OK and EBK services, 61.14 and 11.15 consecutively, were better than two other methods in reliability index. Time index of EBK service was weak (6 min 21 s) but other services were favorable (less that 1 s). Reliability index was impressed under the direct statistical properties (r_RMSE > 0.9). In contrast, time index was more impressionable than structural parameters especially number of sensors (r_IDW=0.99, r_EBK=0.97 and r_OK=0.66). Proposed solution and results would be so useful in environmental monitoring systems development and interaction with other components in IoT context.