Monitoring urban environmental sustainability greening campus using geospatial technology

Abstract

Sustainability is a paradigm which introduces the proper techniques to decrease the harmful effects of human activities, which lead to pollution. Environmental sustainability is at the heart of universities, and the University of Indonesia has introduced a globally accepted sustainability assessment tool called Green metric. This research is focused on Universiti Teknologi Malaysia (UTM) in Johor, which is moving toward achieving environmental sustainability. The physical development of the campus and the increase in its population and transportation have brought about environmental issues like air and noise pollution. In the assessment of land use and transportation sustainability level, it was found that campus physical development is closely related to transportation and is not considered in the Green Metric defining criteria, thus, making it necessary to highlight this issue. As such, the main objective of this study was to develop an approach to address the importance and relationship of urban form and transportation in achieving environmental sustainability. Furthermore, by applying geospatial technology, the indicators could be analyzed and recommended in the Green Metric as an assessment tool, allowing other universities to evaluate their sustainability level. The initial approach used remote sensing data, Worldview 2, and quick birds. Preprocessing included using Dark Pixel, converting the raw digital number value to radiance factor, combining the spectral and spatial resolution attributes of Panchromatic and Multi-Spectral imagery to achieve 0.5 resolution, and using the Gram Schmidt techniques. Next, step-by-step processing actions were applied to the same images using object base classifiers, rule base classifiers and pixel base classifiers, such as maximum like hood and support vector machine, finally, ground truth, result and accuracy assessment were carried out for land cover mapping. Then, Shannon entropy model estimation was calculated to track built-area development, The second approach used interpolation processing techniques. The stages involved were data representing, exploring information, fitting an interpolation model, performing diagnostics, model comparison, and cross-validation using Kriging and Inverse Distance Weight (IDW). The findings of the study area were heterogeneous and complex due to the nature of the urban environment, creating misclassification of the pixels or undesirable information and details. Rule and object classifiers were used to address this issue, and the image object rule base classifier showed 80% overall accuracy, which outperformed the pixel base classified at only 64%. Thus, they are successful techniques for quality and quantity purposes. The estimates found from the Shannon entropy model showed that UTM had become the dividing line between compact and disbursed development, mostly at the T and U zones in the cluster area. Noise, air and traffic pollution in UTM showed medium and high levels, which exceeded the National Ambient Air Quality Standard and noise level threshold. The output for IDW showed good accuracy for nitrogen dioxide and noise pollution levels except for carbon dioxide, which recorded 2.542182 value. Kriging showed good accuracy when the data series was bigger, and a variogram could be calculated. This study has shown that land development and types of traffic movement can be significant parameters for reaching sustainability which was not initially included in the green metrics indicators. As such, multiple type of scenarios related to traffic and land development can be introduced to be helpful for further detection.