Model:DeltaClassification

From CSDMS

DeltaClassification




Metadata

Also known as DeltaClassification
Model type Tool
Model part of larger framework
Incorporated models or components:
Spatial dimensions 2D
Spatial extent Landscape-Scale
Model domain Coastal
One-line model description Geometry classification of delta islands
Extended model description This tool provides a method for extracting information on the nature and spatial extent of active geomorphic processes across deltas from the geometry of islands and the channels around them using machine learning.

The method consists of a two-step ensemble unsupervised machine learning algorithm that clusters islands into spatially continuous zones based on morphological metrics computed on remotely sensed imagery

Keywords:

machine learning, delta landforms,


First name Mariela
Last name Perignon
Type of contact Model developer
Institute / Organization
Postal address 1
Postal address 2
Town / City Boston
Postal code 02101
State Massachusetts
Country United States
Email address mperignon@gmail.com
Phone
Fax


Supported platforms Linux, Mac OS
Other platform
Programming language Python
Other program language Compatible with Python 2.7, not python 3
Code optimized Single Processor
Multiple processors implemented
Nr of distributed processors
Nr of shared processors
Start year development 2017
Does model development still take place? No
If above answer is no, provide end year model development 2019
Code development status As is, no updates are provided
When did you indicate the 'code development status'? 2020
Model availability As code
Source code availability
(Or provide future intension)
Through CSDMS repository
Source web address
Source csdms web address https://github.com/csdms-contrib/DeltaClassification
Program license type BSD or MIT X11
Program license type other
Memory requirements n/a
Typical run time n/a


Describe input parameters Geometric parameters on delta shapes derived from satellite data.

To run this code, the following shape files are required:

• network shapefile, containing the river network extracted from satellite imagery • island shapefile, containing the land masses or islands of the delta • patch shapefile, containing the outline of channels

Input format ASCII
Other input format
Describe output parameters classification of groups of similar zones within a deltasystem

code blocks that:

• loads in the shapefiles • calculate the parameters for the network that both surround and drain the islands • calculate the base metrics (e.g. perimeter, area, solidity, aspect ratio...) • calculates maximum distance from the island center to the nearest water body • estimates minimum, average and maximum widths of all network channels • evaluates the fractal dimension of each delta island • creates shapefiles based on the metrics calculated earlier in the code • saves all metrics to an output file • generates PCA and GeoSOM results from the island and channel metrics • plots the U-matrix and dendrogram based on the GeoSOM results

Output format ASCII
Other output format maps
Pre-processing software needed? No
Describe pre-processing software
Post-processing software needed? No
Describe post-processing software
Visualization software needed? No
If above answer is yes
Other visualization software


Describe processes represented by the model Tool is used to regionalize a study area into zones with 'common physical characteristics' with the underlying aim of differentiating areas of influence of various physical processes. Regionalization attempts to aggregate spatial units or observations into clusters based on spatial continuity as well as attribute similarity.

Geometry metrics are derived from satellite data analysis and include a.o. island area, island aspect ratio, island fractal dimension, and surrounding channel metric, channel width, channel sinousity, number of outflow channels, convexity.

Describe key physical parameters and equations The key methods used are:

1) Feature normalization and principal component analysis 2) Spatial clustering using GEOSOM algorithm 3) Hierarchical agglomerative clustering to built nested clusters

Describe length scale and resolution constraints Methods has been applied to data set of 100's of individual delta islands derived from Landsat satellite data (30-60m resolution).
Describe time scale and resolution constraints Mapview at a given time
Describe any numerical limitations and issues n/a

Limitations on the method do occur when delta systems consist of only a few islands, then the input dataset of geometric parameters becomes too small for the machine learning methods.


Describe available calibration data sets Proof of concept was applied for the Ganges-Brahmaputra delta system
Upload calibration data sets if available:
Describe available test data sets Test data was slightly updated from a published dataset by Passalacqua, P., Lanzoni, S., Paola, C., and Rinaldo, A.: Geomorphic signatures of deltaic processes and vegetation: The Ganges-

Brahmaputra-Jamuna case study, Journal of Geophysical Research: Earth Surface, 118, 1838–1849, 2013.

Upload test data sets if available:
Describe ideal data for testing


Do you have current or future plans for collaborating with other researchers? no
Is there a manual available? No
Upload manual if available:
Model website if any
Model forum / discussion board
Comments Requirements

To run these codes, you will need the following software:

• Python 2.7 or earlier (not compatible with Python 3)

The following Python packages are also required: • matplotlib • scipy • numpy • cPickle • osgeo • fiona • shapely • utilities • sklearn • seaborn • clusterpy • itertools • pandas • pysal • collections

What input is required?

To run this code, the following shape files are required:

• network shapefile, containing the river network extracted from satellite imagery • island shapefile, containing the land masses or islands of the delta • patch shapefile, containing the outline of channels

What does the code do?

The file all.ipynb contains codes run the analysis. From start to finish, the Jupyter Notebook contains code blocks that:

• loads in the shapefiles • calculate the parameters for the network that both surround and drain the islands • calculate the base metrics (e.g. perimeter, area, solidity, aspect ratio...) • calculates maximum distance from the island center to the nearest water body • estimates minimum, average and maximum widths of all network channels • evaluates the fractal dimension of each delta island • creates shapefiles based on the metrics calculated earlier in the code • saves all metrics to an output file • generates PCA and GeoSOM results from the island and channel metrics • plots the U-matrix and dendrogram based on the GeoSOM results


This part will be filled out by CSDMS staff

OpenMI compliant No not possible
BMI compliant No not possible
WMT component No not possible
PyMT component No but possible
DOI model 10.5281/zenodo.3926763
For model version v1.0
Year version submitted 2020
Link to file https://zenodo.org/record/3926763#.Xvz5mpNKiu4
Can be coupled with:
Model info

  • Download DeltaClassification version: v1.0
    Doi: 10.5281/zenodo.3926763
Citation indices DeltaClassification
Nr. of pubs: 2
Citations: 0
h-index: --"--" is not a number.
Qrcode DeltaClassification.png
Link to this page



Introduction

History

References




Citation indices DeltaClassification
Nr. of pubs: 2
Citations: 0
h-index: --"--" is not a number.



Featured publication(s)YearModel describedType of ReferenceCitations
Perignon, Mariela; Adams, Jordan; Overeem, Irina; Passalacqua, Paola; 2020. Dominant process zones in a mixed fluvial-tidal delta are morphologically distinct. .
(View/edit entry)
2020DeltaClassification
Model overview 0
Perignon, M.C.; 2020. csdms-contrib/DeltaClassification: First release of DeltaClassification (Version v1.0). , , . 10.5281/zenodo.3926763
(View/edit entry)
2020DeltaClassification
Source code ref. --
See more publications of DeltaClassification


Issues

Help

Input Files

Output Files