Abstract
Correction of digital elevation models (DEMs) for flat areas is a critical process for hydrological analyses and modeling, such as the determination of flow directions and accumulations, and the delineation of drainage networks and sub-basins. In this study, a new algorithm is proposed for flat correction/removal. It uses the puddle delineation (PD) program to identify depressions (including their centers and overflow/spilling thresholds), compute topographic characteristics, and further fill the depressions. Three different levels of elevation increments are used for flat correction. The first and second level of increments create flows toward the thresholds and centers of the filled depressions or flats, while the third level of small random increments is introduced to cope with multiple threshold conditions. A set of artificial surfaces and two real-world landscapes were selected to test the new algorithm. The results showed that the proposed method was not limited by the shapes, the number of thresholds, and the surrounding topographic conditions of flat areas. Compared with the traditional methods, the new algorithm simplified the flat correction procedure and reduced the final elevation increments by 5.71–33.33%. This can be used to effectively remove/correct topographic flats and create flat-free DEMs.
Highlights
Digital elevation models (DEMs) have been widely used for hydrological analyses and modeling
To eliminate the stagnation of water flowing from depressions to stream networks, surface depressions need to be fully filled before calculating flow directions and accumulations
Such small flats may seriously influence the efficiency of their surface delineation method [2,15,30], in which the search process starts from sinks and flat areas
Summary
Digital elevation models (DEMs) have been widely used for hydrological analyses and modeling. Zhang et al [17] overlaid a digital channel network on flat areas, and used an interpolation method to make water flow toward channels. Barnes et al [29] made some improvement with FlatMask by applying different weight gradients in the delineation processes, their approach tended to create more new small flat areas. Such small flats may seriously influence the efficiency of their surface delineation method [2,15,30], in which the search process starts from sinks and flat areas. A new algorithm is proposed to correct/remove flat areas by adding small elevation increments for all flat cells. The new algorithm is tested by using a set of artificial surfaces and two real-world surfaces
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Free) 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
