gis

In April of this year, I made an animation of the 24-hour average temperature of January 2020, also showing the day-night cycle. My biggest problem was finding a way to project correctly the area at night without breaking the geometry. The easiest solution I found was rasterising the night polygon and then reprojecting it. Indeed, a vector approach could be used, but I have preferred to use raster data here.

A disadvantage of choropleth maps is that they tend to distort the relationship between the true underlying geography and the represented variable. It is because the administrative divisions do not usually coincide with the geographical reality where people live. Besides, large areas appear to have a weight that they do not really have because of sparsely populated regions. To better reflect reality, more realistic population distributions are used, such as land use. With Geographic Information Systems techniques, it is possible to redistribute the variable of interest as a function of a variable with a smaller spatial unit.

I recently created a visualization of the distribution of river flow directions and also of coastal orientations. Following its publication in social networks, I was asked to make a post about how I did it. Well, here we go to start with an example of rivers, coastal orientation is somewhat more complex.

The first post of this year 2020, I will dedicate to a question that I was recently asked. The question was how to calculate the shortest distance between different points and how to know which is the closest point. When we work with spatial data in R, currently the easiest thing is to use the sf package in combination with the tidyverse collection of packages. We also use the units package which is very useful for working with units of measurement.