Foreword
This interactive module has been developed with the R programming and statistical language and Rstudio IDE. It has been made possible with a great number of additional packages developed by the R community, most importantly knitr, rmarkdown, bookdown, shiny and learnr. They are amazing packages for integrating R code and formatted text together in documents and webpages that you will hopefully find good looking and functional!
The data analysis itself can be carried out with a number of applications, but R has been chosen because it’s free, opensource, has a large and active community of users that can help each others. It has solutions for all the steps required to take data from a forest inventory field campaign and derive statistics.
The formulas and concepts applied here were taken from the NFI training modules [LINK] and this interactive module is meant to complete the theoretical NFI training modules with some practice on sampling design and deriving inference for different population variables of interest.
If you don’t have any experience of programming languages and of R in particular, we strongly recommend you to go through a general basic introduction to R before you complete this module. For an introduction to R there are plenty of resources freely available on the internet but if you don’t know where to start, and since a good amount of our code is based on the tidyverse collection of R packages, you could try R for Data Science. The book introduction will guide you through installing R, R studio, key packages and the main functions that you use here for data wrangling.
A fair amount of code is related to spatial analysis and spatial data visualization. The online book [Geocomputation with R]{https://geocompr.robinlovelace.net/} is a good place to get more information and a deeper unerstanding on some of the functions used here.
Version’s notes
V1.0
- One country based on a 90 km square with a 30 m resolution topogr4phy and land cover map.
- Five forest types including mangrove forest.
- Random and systematic sampling design.
- Fixed 20 circular plot designs
- Calculation chain for the average forest aboveground carbon stock
Stay tuned for V2.0. New features planned:
- Choose between several land profiles.
- Customize the number of forest categories and climatic conditions.
- Practice stratified sampling design.
- Practice different plot and cluster-plot designs.
- Estimate additional indicators such as biodiversity.
0.1 R packages list
Packages that we will actively use:
tidyversesf
Packages used for developing the tutorial:
knitrrmarkdownbookdownshinylearnrtmap