njurs / sarn-method

Code for river connection

SARN (Sentinel-ArcticDEM merged River Network)

Environment

Python with ArcPy package

Prepare data

• DEM-modeled drainage network.shp: the DEM-modeled drainage network polyline
• RS-mapped river network.shp: the river centerlines of the RS water body mask. We used the river detection method proposed by Kang Yang (https://github.com/njuRS/River_detection.git). You can use the customed RS water body mask instead.
• RS water mask.tif: the binary water body mask, water is set to 1, non-water is 0

In order to ensure this code runs correctly, please make sure the coordinates are consistent

Create Geodatabase

Create a File Geodatabase.gdb and import the three files in it. All process data will be saved in this database.

Run river connection code

• Open connect_river.py in Python.
• Set input 6 parameters.

parameters	description
workspace	the path of the geodatabaseset
DEM	the name of the DEM-modeled drainage network
RS_vector	the name of the RS-mapped river centerlines
RS_Raster	the name of the RS binary water body mask
cell size	the cell size of the images, Sentinel-2 imagery is 10
delete_length	river length less than 300 m will be deleted

• Click run.

• These two shapefiles in the red box are the result. unspit_split_coincide + DEM name.shp is the RS-mapped river segments, and unspit_split_connect + DEM name.shp is the DEM-modeled river segments.

Acknowledgements

Cite

Lu, X., Yang, K., Lu, Y., Gleason, C.J., Smith, L.C., & Li, M. (2020). Small Arctic rivers mapped from Sentinel-2 satellite imagery and ArcticDEM. Journal of Hydrology, 584, 124689

Corresponding Author

• Xin Lu ([email protected], ph:15951756762, School of Geography and Ocean Science, Nanjing University)
• Kang Yang ([email protected], School of Geography and Ocean Science, Nanjing University)
• Yao Lu ([email protected], School of Geography and Ocean Science, Nanjing University)