METHOD FOR SHALLOW DRAINAGE DITCH NETWORK GENERATION USING REMOTE SENSING DATA

Abstract

Aim of this study is to develop a method for automatic shallow drainage ditch generation to drain terrain depressions using four factor least cost surface which is obtained using LiDAR (light detection and ranging) data and Sentinel-2 multispectral satellite imagery. LiDAR data are used for depression mapping in DEM, flow accumulation and slope modelling as well as CHM (canopy height model) to obtain relative vegetation height. Sentinel-2 imagery was used for land cover type identification as well as separating coniferous and deciduous forest stands. Study area is located in western Latvia and is 25 km² large. Least cost surface connects DEM depressions and already existing drainage ditches by best possible path for shallow ditch network digging. Different methods are applied to determine depressions which can be drained as well as changes of affected drained area and depression depth. This results in suitable areas where to create shallow ditches to improve water runoff. Results show that using this method average reduction of area of depressions is 79% and average length of shallow ditches on each drained depression hectare is 370 m.