Vahidi Hossein
Despite the considerable benefits of using of Volunteered Geographic Information (VGI) for enriching and updating the tree inventory databases, the skepticism regarding the data quality remained as one of the main barriers to deploy of VGI for creating the authoritative tree inventories. Therefore, developing the effective and robust VGI quality assurance procedures for verifying and improvement of the VGI quality in the tree inventory databases may help to overcome the existing concerns regarding to the VGI fitness of use for the serious scientific and professional applications. This thesis addressed the different open problems in the area of the data quality assurance in the crowdsourced tree inventory databases.
Firstly, this research focused on the quality assurance of VGI that has been collected across the globe for Citizen Science (CS) tree species diversity monitoring programs. The quality assessment of the obtained crowdsourced tree species occurrences in these program via cross-checking of them against the reference dataset is almost an impossible task as acquisition of the detailed authoritative species occurrence data over a large area is very costly and time consuming. For this end, a novel intrinsic VGI quality indicator and a fuzzy model for quality assessment of the crowdsourced tree species occurrence observations was developed in this research that reduces the dependence of VGI quality assurance processes on authoritative data, and the contribution of experts or the community. The proposed approach evaluates the thematic and positional quality of the crowdsourced tree species observations in terms of trustworthiness of the VGI by combining three indicators of consistency with habitat, consistency with surroundings, and reputation of contributor, that characterize the geographical and social aspects of trust to VGI. To evaluate the performance and usability of the proposed approach for evaluating the trustworthiness of crowdsourced observations and detecting thematic and positional errors in crowdsourced observations, the developed approach was applied to the crowdsourced observations on Acer macrophyllum. The result of a conformity test at the optimal acceptance threshold (sensitivity=0.99, specificity=0.8, and Cohen’s kappa=0.79), the achieved area under the curve (AUC) value (AUC=0.98), and the results of the complementary investigation on the predictions of the proposed model indicated that the proposed fuzzy trust model exhibited promising predictive performance and was able to flag the majority of attribute and positional errors amongst the observations in the crowdsourced tree species inventories
Second, this research concentrated on the improvement of the VGI quality in the tree inventory databases that were established for inventory of trees in private urban orchards. The low completeness degree of the obtained VGI in urban orchard tree inventory databases makes it difficult to use this data for means of the formal monitoring and conservation applications. To fill the existing gap, a new approach for improvement of the completeness degree of remotely VGI (i.e., the VGI that was obtained via remote mapping –a.k.a. armchair mapping– technique) on urban orchard trees was proposed in this study. For this end, an automatic tree detection workflow based on the collective sensing approach was developed to enhance the initial poor completeness degree of the gathered VGI via remote mapping technique (C = 22.5% for the study area) in order to establish a high quality database for inventorying of the trees in urban orchards. In the proposed workflow, the detection of the orchards trees in the very high-resolution (VHR) optical satellite imagery was performed by adopting a Template Matching (TM) approach that deploys the available VGI in the tree inventory database for creating the required templates. The assessment of the quality of the detected tree features in this study demonstrated the usefulness and effectiveness of the proposed approach for creating a reliable dataset on urban orchard trees as the proposed approach could dramatically improve the initial completeness degree of the data by achieving a very high completeness degree (C = 92.7% for the study area) while it was maintaining the thematic and positional quality of the data at a high standard level (F_1= 0.918 and RMSE = 1.02 m)
Finally, the thesis proposed a holistic conceptual model of a system for quality assurance of the generated crowdsourced occurrence data in the collaborative platforms for the crowdsourced tree species inventory. The different possible ex-post and ex-ante architectures for creating a quality assurance system for validating and improving the quality of the generated VGI in the collaborative platforms for the crowdsourced tree species inventory. The proposed conceptual models empower the end user (data consumer), expert reviewers, and volunteers (data producers) to perform more robust and precise VGI quality assurance practices.
This study shed a light on issue of the VGI quality in the tree inventory databases for the first time. A set of robust and efficient approaches for assuring the quality of obtained VGI in the different types of crowdsourced tree inventory databases was developed in this thesis. In conclusion, the exhibited promising performance of the proposed approaches indicated that the suggested rigorous methodologies in this study have a high potential to be adopted and embedded systematically in the crowdsourced tree inventory databases. In this sense, the deployment of the proposed approaches may address a number of the most critical open challenges in the area of quality assurance of the observations in these databases and ensure the stakeholder to rely on these databases in the formal and serious applications.
- Vahidi, H., & Yan, W. (2014). Towards Spatially Explicit Agent-based Model for Simulation of Informal Transport Infrastructure Indirect Growth Dynamic in Informal Settlements. The International Archives of the Photogrammetry, Remote Sensing and Spatial Information Sciences, Volume XL-2/W3, 2014 The 1st ISPRS International Conference on Geospatial Information Research, 15–17 November 2014, Tehran, Iran, XL(November), 15–17. https://doi.org/10.5194/isprsarchives-XL-2-W3-273-2014
- Vahidi, H., & Yan, W. (2016). How is an informal transport infrastructure system formed ? Towards a spatially explicit conceptual model. Open Geospatial Data, Software and Standards, 1–26. https://doi.org/10.1186/s40965-016-0009-9
-
Pouriyeh, A., Khorasani, N., Hosseinzadeh Lotfi, F., & Farshchi, P. (2016). Efficiency evaluation of urban development in Yazd City, Central Iran using data envelopment analysis. Environmental Monitoring and Assessment, 188(11). https://doi.org/10.1007/s10661-016-5548-0
-
Vahidi, H., Klinkenberg, B., & Yan, W. (2016). An Interactive System for Intrinsic Validation of Citizen Science Data for Species Distribution Mapping and Modelling Applications. LBS 2016, November, 14–16
-
Vahidi, H., Klinkenberg, B., & Yan, W. (2018). Trust as a proxy indicator for intrinsic quality of Volunteered Geographic Information in biodiversity monitoring programs. GIScience and Remote Sensing, 55(4), 502–538. https://doi.org/10.1080/15481603.2017.1413794
-
Vahidi, H., Klinkenberg, B., Johnson, B. A., Moskal, L. M., & Yan, W. (2018). Mapping the Individual Trees in Urban Orchards by Incorporating Volunteered Geographic Information and Very High Resolution Optical Remotely Sensed Data : A Template Matching-Based Approach. Remote Sensing, 10(online), 1–42. https://doi.org/10.3390/rs10071134
