Water bodies face continuous pressure from human activities, particularly from nitrogen and phosphorus from agricultural activities and urban wastewater discharges, leading to water pollution. Studies have shown that monitoring water quality using the traditional method of water sampling is time-consuming, labour-intensive, costly and cannot be done frequently. In environmentally sensitive areas, these techniques can encounter additional restrictions in relation to access, team size, etc. Studies have indicated that although satellite-based imagery has clear potential as a non-contact methodology, it has limitations due to the coarse spatial resolution and potential for the site to be obscured due to cloud cover. Additionally, while commercial or airborne imagery can provide high spatial resolution, they are often expensive to use. Drones have shown the potential to provide very high spatial resolution at regular intervals and at a lower cost, acting as a middle ground between satellite, manned aircraft, and terrestrial insitu methods. However, issues concerned when mapping water bodies using lowcost drones are largely unknown. To date, no methods involving low-cost drones have been investigated to (i) understand their suitability, (ii) the associated errors have never been quantified, and (iii) there have been no attempts to improve accuracy. To undertake this investigation, a theoretical analysis was first carried out to benchmark the accuracy of the Direct Georeferencing (DG) method for mapping water bodies and validate with real-world data. Following this, the errors associated with the DG method were identified and quantified, resulting in the development of photogrammetric methods to improve accuracy by correcting for these errors. The first method developed an analytical photogrammetric approach to improve the accuracy of drone imagery over water. The second method was an advanced approach that exploited the shore to enhance the accuracy of drone imagery when mapping water bodies. The final approach developed a scalable method that could improve the accuracy and map water bodies at different geographic scales. The developed methods presented a novel approach to improve the accuracy of drone imagery for mapping water bodies. Institutions, organisations and researchers can utilise the methods developed in this thesis to generate accurate drone imagery of water bodies which can be used to measure various water quality parameters.