Skip to content
I have found that one of the biggest challenges as a scientist working on global change issues in South Africa has been not having access to my country’s long-term weather records. Our weather service charges for the distribution of data (despite it being collected with taxpayer’s money), and a hefty fee at that.
While one can occassionally access data for research purposes, this is usually only if there’s a student involved, and requires tracking down an individual within the organization who’s willing to field your query.
The Basics Interlude Doing GIS from R Since I first started maintaining blog posts on handling spatial data in R perhaps the most common question I’ve received is “How do I handle big spatial data in R?”. I thought its finally time to provide a blog post to deal with this particular topic. The answer of course is that there are many, many ways.
Now I know those of you who have asked me in person are thinking “That’s not what he said when I asked?
A quick note on the structure of this tutorial Data Description Housekeeping Getting and cleaning the data, but first and foremost, projection!!! Let’s start with point data Raster data (mostly functions from library(raster)) Polygons! Going parallel!!! Animation! Some other data visualization and analysis But what about our poor cedars? Options for writing out spatial data This post follows on from Handling Spatial Data in R - #1.
Installing R (and RStudio) CRAN Spatial Task Views!!! Useful DIY resources Points, lines, polygons and rasters - R can handle them all. My aim for this post is to give you the basics required to teach yourself spatial data analysis in R - following 3 major sections listed above.
In the next post I provide a practical example working with point, line, polygon and raster data. If you’re already familiar with R then you can skip straight there, but you may be interested to check out the useful DIY resources before you go.
This is based on data collected when I took students from SAEON’s Graduate Student Network into the field to introduce them to Fynbos and to show them a bit of the science I’m working on. To make things more interesting, I got them to help me test a smartphone app I’ve been developing (“VeldWatch”) that can be used to map any impacts on natural ecosystems observed in the field, like invasive species, plant mortality, etc.
This is a study recently completed by one of my MSc students, Annabelle Rogers, who I co-supervised with Prof Ed February and Glenn Moncrieff. We’re busy writing it up for publication, but here’s a little preview.
The study was inspired by this great paper published by Dr Coert Geldenhuys in 1994, explaining how fire determines the distribution of forests in the Southern Cape. You can also read about it, and get a nice intro to our forests, in this interesting SA Forestry Online article.
This is a quick primer on how to handle MODIS and Landsat NDVI data from Google Earth Engine (GEE) in R. It’s primarily for my Honours student, Hannah Simon, but I thought why not make it public for all to share? I first posted the MODIS section and had Hannah play with the Landsat data as a learning exercise, so full credit to her for the Landsat code.
For a general introduction for many of the functions used and handling spatial data in general see A Primer for Handling Spatial Data in R posted previously.