Digital Elevation Model (DEM) and KnickZone Picker (KZP) processing
Code developed by Al Neely and Bodo Bookhagen 10/12/2015, significantly modified March-May 2016, and May 2017
The code has been tested with Matlab R2012b, R2014b and R2015b, R2016b, R2017a. It requires the Statistical Toolbox, the Topotoolbox, and export_fig (see below). If the Curve Fitting Toolbox is available, it will be used (smooth). If no Curve Fitting Toolbox is available, sgolayfilt is used (with similar and comparable results).
Installation
Before running the code and taking advantage of all feautres (i.e., high-resolution figures and projected shapefile), you will need to install the additional items:
-
TopoToolbox (https://topotoolbox.wordpress.com/download/). I suggest using the github repository (https://github.com/wschwanghart/topotoolbox). For additional information, see: https://topotoolbox.wordpress.com/. The TopoToolbox requires the Image Processing Toolbox and the Statistics Toolbox (which are mostly installed in academic environment). The Mapping Toolbox will come in handy, but is not required. The Curve Fitting Toolbox will be used if available and will produce slightly different fitting parameters, e.g. (for 95% confidence Intervals). Make sure to add the Topotoolbox to the Matlab PATH.
-
export_fig from Mathworks MATLAB Central: (http://www.mathworks.com/matlabcentral/fileexchange/23629-export-fig) for creating the various plots and figures in high-quality PDF format. Make sure to add the location of the file export_fig.m to the Matlab PATH.
-
Install ghostscript from http://www.ghostscript.com/download/gsdnld.html . This is required for high-res PDF outputs for export_fig.m and highly recommended.
-
OSGEO4WShell or the full GDAL suite (http://www.gdal.org/ or http://trac.osgeo.org/osgeo4w/). MAC OS X users should install the GDAL complete framwork from http://www.kyngchaos.com/software/frameworks. This will allow to generate shapefiles from
tables that are created during the processing and will allow to polygonize raster (TIF) files. GDAL is freely available for all operating systems, and we have tested this on Ubuntu 12.04 (sudo apt-get install gdal), Windows 7 and 8.1 (http://trac.osgeo.org/osgeo4w/), and Mac OS X (http://www.kyngchaos.com/software/frameworks). This script will run well, if you use the default gdal installation options. Otherwise, you may have to change the directory locations in the parameter file. We rely on: ogr2ogr, gdalsrsinfo, gdal_dem, gdal_polygonize. -
Install the KnickZonePicker (KZP) code and subdirectories and add the KZP directory to the Matlab PATH. You can use addpath or the GUI.
This Matlab code takes MAT file written by KZP_topometrics_v1.m and calculates knickpoints and additional attributes from chi-plot analysis. It will write several shapefiles and csv files that can be read into any GIS software.
You will have to run knickpoints_preprocessing_v1.m at least once before running this script.
Parameters for this script can be adjusted/changed in knickpoints_parameters_X.m.
This file will output the following four shapefiles:
1: _kp_bases.shp (knickpoint bases for all tributaries)
2: _kp_lips.shp (knickpoint lips for all tributaries)
3: _kp_bases_trunk.shp (knickpoint bases for trunk stream)
4: _kp_lips_trunk.shp (knickpoint lips for trunk stream)
The code developes a database/csv file/shapefile for the bottom and top points of all knickzones. Each database has the following attributes with their respective attribute name in brackets [].
Knickpoint Database information [shapefile attribute]
1: knickpoint # [1kp_id]
2: stream id # [2stream_id]
3: tributary id # [3trib_id]
4: slope of stream, bfl slope: elev/chi [4sl_str]
5: chi coordinate [5chi]
6: elevation (m) [6elev_m]
7: knickpoint magnitude, detrended elevation drop (m) [7kp_magnt]
8: Knickpoint Relief, height of knickpoint (not detrended (m)) [8kp_Rel]
9: easting (meters utm) [9Easting_m]
10: northing (meters utm) [10North_m]
11: upstream drainage area (m2) [11DA_m2]
12: dist upstream (m) [12kp_DFM]
13: knickpoint slope (elev drop/distance upstream) [13kp_slp]
14: elevation drop (detrended elevation drop/distance upstream, m) [14kp_slp_d]
15: knickpoint slope (detrended elevation drop/chi) [15kp_slp_dt/c]
16: knickpoint length (distance usptream, m) [16kp_len_m]
17: sgolay smoothing window size (grid cells) [17sgol_smv]
18: knickpoint lumping search window size (grid cells) [18kp_lm_ws]
19: minimum knickkpoint size pre-lumping [19kp_pr_lu]
20: minimum knickpoint size post-lumping (final minimum knickpoitn size) [20kp_pt_lu]
21: minimum steepness anomaly [21kp_stp_a]
22: minimum stream size for analysis (cells) [22strea_sz]
#In order to process a DEM and identify Knickpoints, follow these steps: (1) Load DEM file and other data from preprocessing KZP_processing_1load_v1
(2) Iterate through all basin tributaries and extract knickpoints KZP_processing_2kpp_tribs
(3) Iterate through all trunk streams and pull out knickpoints KZP_processing_3kpp_trunks
(4) Generate figures for each basin KZP_processing_4kpp_mkfigs