Use of Bathymetric and LiDAR Data in Generating Digital Elevation Model over the Lower Athabasca River Watershed in Alberta, Canada

• Published in: Water
• Main Authors: Ehsan H. Chowdhury, Quazi K. Hassan, Gopal Achari, Anil Gupta
• Format: Article in Journal/Newspaper
• Language: English
• Published: MDPI AG 2017
• Subjects: cross section; geoswath; inverse distance weighting; kriging; predicted surface; Hydraulic engineering; TC1-978; Water supply for domestic and industrial purposes; TD201-500; Fort McMurray; Athabasca River; Canada; Firebag River
• Online Access: https://doi.org/10.3390/w9010019; https://doaj.org/article/2c8e47199dcb42af88c42f714f6172a7

The lower Athabasca River watershed is one of the most important regions for Alberta and elsewhere due to fact that it counts for the third largest oil reserve in the world. In order to support the oil and gas extraction, Athabasca River provides most of the required water supply. Thus, it is critical to understand the characteristics of the river and its watershed in order to develop sustainable water management strategies. Here, our main objective was to develop a digital elevation model (DEM) over the lower Athabasca River watershed including the main river channel of Athabasca River (i.e., approximately 128 km from Fort McMurray to Firebag River confluence). In this study, the primary data were obtained from the Alberta Environmental Monitoring, Evaluation and Reporting Agency. Those were: (i) Geoswath bathymetry at 5–10 m spatial resolution; (ii) point cloud LiDAR data; and (iii) river cross-section survey data. Here, we applied spatial interpolation methods like inverse distance weighting (IDW) and ordinary kriging (OK) to generate the bathymetric surface at 5 m × 5 m spatial resolution using the Geoswath bathymetry data points. We artificially created data gaps in 24 sections each in the range of 100 to 400 m along the river and further investigated the performance of the methods based on statistical analysis. We observed that the DEM generated using the both IDW and OK methods were quite similar, i.e., r2, relative error, and root mean square error were approximately 0.99, 0.002, and 0.104 m, respectively. We also evaluated the performance of both methods over individual sections of interest; and overall deviation was found to be within ±2.0 m while approximately 96.5% of the data fell within ±0.25 m. Finally, we combined the Geoswath-derived DEM and LiDAR-derived DEM in generating the final DEM over the lower Athabasca River watershed at 5 m × 5 m resolution.