Procesamiento de series de tiempo en GRASS GIS: Aplicaciones en Ecologia y Ambiente.

Edit: Adopted for Coding in Geo Kenya Mini Webinar Session for demo and practice purposes 2021

Data, code and slides for the post-graduate course that will be held in Rio Cuarto (Cordoba, Argentina) in October, 2018

Slides and exercises

Slides:

• GRASS GIS intro
• GRASS GIS general capabilities
• Raster data processing
• Satellite imagery processing
• Temporal data processing
• GRASS and R interface

Exercises:

• Getting familiar with GRASS GIS
• Create new location and import maps
• Landscape, hydrology and terrain analysis
• Working with Sentinel 2 images
• Hands-on to NDVI time series
• GRASS and R: Predicting species distribution

Software

We will use GRASS GIS 7.8.5 (current stable version). It can be installed either
through standalone installers/binaries or through OSGeo-Live (which includes all
OSGeo software and packages).

Standalone installers for different OS:

MS Windows

There are two different options:

1. Standalone installer: 32-bit version | 64-bit version
2. OSGeo4W package (network installer): 32-bit version | 64-bit version

For Windows users, we strongly recommend installing GRASS GIS through the OSGeo4W package,
since it allows to install all OSGeo software. If you choose this option,
make sure you select GRASS GIS and msys. The latter one will allow
the use of loops, back ticks, autocomplete, history and other nice bash shell
features.

Mac OS

Download GRASS GIS 7.8.5 stable from:http://grassmac.wikidot.com/downloads and follow the instructions under Installing GRASS for Mac.

Ubuntu Linux

Install GRASS GIS 7.8.5 from the “unstable” package repository:

Note: packaged with GDAL 3 or later
If you already have QGIS installed through official docs with qgis-grass-plugin then you probably have grass installed. QGIS Grass GIS

sudo add-apt-repository ppa:ubuntugis/ubuntugis-unstable
sudo apt-get update
sudo apt-get install grass

Fedora, openSuSe Linux

For other Linux distributions including Arch Linux, EPEL, Fedora and openSuSe, simply install GRASS GIS with the respective package manager. See also here

Docker users

If you’re comfortable with using Docker, check instructions here

Extra dependencies

The following are some Python libraries that are needed by add-ons that will be used in the course:

• pyModis
• sentinelsat
• pandas
• matplotlib
• pygbif

See the Installation guide presentation for details.

OSGeo-live:

OSGeo-live is a self-contained bootable DVD, USB thumb
drive or Virtual Machine based on Lubuntu, that allows you to try a wide variety
of open source geospatial software without installing anything. There are
different options to run OSGeo-live:

• Run OSGeo-live in a Virtual Machine
• Run OSGeo-live from a bootable USB flash drive

For a quick-start guide, see: https://live.osgeo.org/en/quickstart/osgeolive_quickstart.html

GRASS GIS Add-ons that will be used during the course

• r.diversity: Calculates diversity indices based on a moving window
• r.forestfrag: Computes the forest fragmentation index
• r.stream.distance: Calculates distance to and elevation above streams and outlet
• r.lake.series: Fills lake at given point(s) to given levels
• i.sentinel: Toolset for download and processing of Copernicus Sentinel products
• i.fusion.hpf: Fusion of high resolution panchromatic and low resolution multi-spectral data based on the High-Pass Filter Addition technique
• i.landsat8.qc: Reclassifies Landsat8 QA band according to pixel quality
• i.wi: Calculates different types of water indices
• i.superpixels.slic: Perform image segmentation using the SLIC segmentation method
• i.modis: Toolset for download and processing of MODIS products using pyModis
• r.hants: Approximates a periodic time series and creates approximated output
• r.seasons: Extracts seasons from a time series
• r.regression.series: Calculates linear regression parameters between two time series
• v.strds.stats: Zonal statistics from given space-time raster datasets based on a polygons vector map
• v.in.pygbif: Search and import GBIF species distribution data with filters
  <!—-
• r.learn.ml: Supervised classification and regression of GRASS GIS raster maps using the python scikit-learn package
  —->
  Install with g.extension extension=name_of_addon

Data

• North Carolina location (full dataset, 150Mb): download and unzip within $HOME/grassdata.
• Sample raster maps: download and unzip within $HOME/gisdata.
• Streets vector map (gpkg): download and move into $HOME/gisdata.
• Landsat 8 scenes clipped to North Carolina: download and unzip within $HOME/gisdata.
• Sentinel 2 scene (500Mb): download and move into $HOME/gisdata. DO NOT UNZIP.
• Aeronet AOD file: download and unzip within $HOME/gisdata.
• modis_lst mapset (2Mb): download and unzip within the North Carolina location in $HOME/grassdata/nc_spm_08_grass7.
• modis_ndvi (15 Mb): download and unzip within the North Carolina location in $HOME/grassdata/nc_spm_08_grass7.

The trainer

Verónica Andreo is a researcher for CONICET
working at the Institute of Tropical Medicine (INMeT)
in Puerto Iguazú, Argentina. Her main interests are remote sensing and GIS tools
for disease ecology research fields and applications.
Vero is an OSGeo Charter member and a FOSS4G
enthusiast and advocate.
She is part of the GRASS GIS Development team
and she also teaches introductory and advanced courses and workshops, especially
on GRASS GIS time series modules
and their applications.

Contributors

Many thanks to GRASS developers and community members that have developed other
educational materials from which I recycled and adapted some of the examples for
this course. A special thanks to Carol Garzon
who contributed the examples for r.li.* modules and the workflow for species
distribution modeling in R that were further adapted for this course.

Post questions or discussions here

License

All the course material:

Creative Commons Attribution-ShareAlike 4.0 International License

Presentations were created with gitpitch:

• MIT License