A GIS-based Land Cover Classification Approach Suitable for Fine‐scale Urban Water Management

• Published in: Water resources management Vol. 35; no. 4; pp. 1339 - 1352
• Main Authors: Hiscock, Oscar H., Back, Yannick, Kleidorfer, Manfred, Urich, Christian
• Format: Journal Article
• Language: English
• Published: Dordrecht Springer Netherlands 01.03.2021; Springer Nature B.V
• Subjects: Accuracy; Atmospheric Sciences; Austria; Case studies; Civil Engineering; Classification; climate; computer software; concrete; Earth and Environmental Science; Earth Sciences; Environment; Geographical information systems; geometry; Geotechnical Engineering & Applied Earth Sciences; Grasses; Hydrogeology; Hydrology/Water Resources; Imagery; infrastructure; irrigation; Land cover; Management analysis; Methodology; Population growth; radiometry; Urban areas; Urban planning; Urbanization; water; Water engineering; Water management; Water supply systems; Austria; GIS; Land cover classification; Multi‐functional; Urban water Management; NDVI
• ISSN: 0920-4741; 1573-1650
• DOI: 10.1007/s11269-021-02790-x

In the context of climate stress, urbanisation and population growth, design and planning tools that assist in decentralised and environmental infrastructural planning are becoming more common. In order to support the design of increasingly complex urban water infrastructure systems; accurate and easily obtainable spatial databases describing land cover types are crucial. Accordingly, a methodology categorizing land covers that supplements these tools is proposed. Utilizing GIS imagery of high spatial accuracy that is easily obtainable from flyover techniques, radiometric and geometric data is generated to create a multi-functional classification of urban land cover, designed to be applicable to various urban planning tools serving different purposes, e.g. urban water management. The methodology develops 13 individual land cover categories based on the complete capabilities of the NDVI and nDSM imagery, which is then adapted to suit planning tool requirements. Validation via a case study application at Innsbruck (Austria), an overall classification accuracy of 89.3 % was achieved. The accuracy of the process was limited in differentiating certain categories (e.g. Dry Grass and Concrete, Trees and Irrigated Grass, etc.), which could yield limitations subject to intended model applications. Despite this, the classification results yielded high accuracy, demonstrating the methodology can be utilised by various software to improve urban water management analysis.