Methods & Data

Methods

The area of study is the continent of Africa lying between 37.35°N and 34.85°S and; 17.52°W and 51.46°E. This study categorized the continent into seven climatic zones; Bushveld, Desert, Equatorial, Mediterranean, Savanna, Steppe, and Sub-tropical (Fig 1). The shapefile for this map was customized from the base map of Agroecological zones of Africa of climate, moisture and elevation dimensions (Sebastian, 2009).

The zones used in this map and study were based on categorization by the Köppen-Geiger climate classes (Beck et al., 2018), the climatological study on equatorial Africa (Donahue et al., 1995) and a publication on the Climate of Africa by Britannica.

Fig 1: The map was customized to this study using the base map of Africa's Afro-ecological zones (Sebastian, 2009)

Inset map showing Africa on the World Map. Developed from QGIS embedded OpenStreetMap.

The climatic zones

  1. Bushveld: Open country in the southwestern parts of the continent covered by pasture and farmland.

  2. Desert: Spanning across the northern areas of the continent, the horn in the east and parts of southeastern Africa. Covered by bare dunes of sand and desert biota.

  3. Equatorial: Also the tropical wet covering middle and western Africa along the equator (Latitude 0°) and the western scape of Madagascar. Dominated by rainforest.

  4. Mediterranean: Covering the northernmost and southernmost parts of Africa typically covered with grasslands.

  5. Savanna: arid lands transitioning from desert towards the sub-tropical covered by woodlands or grasslands.

  6. Steppe: arid lands in the southeastern parts of the continent covered by vast grasslands.

  7. Sub-tropical: Typically around the tropics and mid-madagascar covered by deciduous or moist forest, grasslands and farmlands.

Fig 2: Geographical location of meteorological stations from which data for this project was collected.

The climatic data was collected from meteorological stations located on the continent (Fig 2). Monthly data was used to calculate the mean annual temperature (°C) and the total annual precipitation (mm). The general period covered for this project was 1900 to 2022 for temperature and 1900 to 2010 for precipitation.

The data sets used were obtained from:

The data was imported into R and checked for reliability. A visual analysis was done to show the recordings for each station. There were visibly fewer recordings for the period 1900 to 1950 than in the latter years. Fig 3 represents all the MAT station data for the Sub-tropical zone (1900-2022) and after it was filtered. Some stations also had few recordings over the years. All the data from 1900 was still used in visual analyses (exploratory graphs) with predictions using regression linear models to fill the gaps and increase statistical power however only filtered data (1950 onwards) and the with 30 or more recordings (N>=30) was used for statistical analyses.

Fig 3 Recordings for the Mean Annual Temperature (°C) (MAT) for stations in Sub-tropical Africa (1900-2022) and (1950-2022)

Since the meteorological station data was not categorized by the climatic zones, it was imported in QGIS and clipped with the climatic zonal layers, and merged in R. This was then cleaned and reviewed, and analyzed in R, Microsoft Office Excel and PowerPoint.

This data was limited by unequal availability and somewhat insufficient weater stations data across the regions for example the northern region generally had fewer stations for both temperature and precipitation. The climatic zones used did not account for local climates as those around mountain areas and water bodies.

Data

Data collected was used to analyze two climatic variables; the mean annual temperature (°C) (MAT) and total annual precipitation (mm) (TAP). Review Table 1 which is an abbreviated table of the data used in Total Annual Precipitation (TAP) analysis. Monthly data values (greyscale) were aggregated (averages for temperature and sum for precipitation) to obtain a single value (TAP) per station (orange column) per year(yellow column). The year and climatic zone "CLIMZONE" were the predictor variables and the recorded values for temperature and precipitation were the response variables. Whereas the values for the. The response variables were also influenced by the climatic zone "CLIMZONE" (green column).

Table 1: Data table used in the analysis of Total Annual Precipitation (TAP) of the different climatic zonesPrimary data reformed to this table was downloaded from http://doi.org/10.5281/zenodo.3520885

The data for MAT and TAP was diagnosed using boxplots in R where outliers were observed as in Fig 4(a) for MAT and Fig 4(c) for TAP. These were cleaned out using Tukey's Inner Fence (1.5*(Q3-Q1)) where Q3 is the upper quartile and Q1 is the lower quartile Fig 4(b) and Fig 4(d).

Fig4(a) Recordings for the Mean Annual Temperature (°C) (MAT) for stations in Eastern Africa (1900-2022)

Fig 4(b) Recordings for the Mean Annual Temperature (°C) (MAT) for stations in Eastern Africa (1950-2022) that had more than 20 annual recordings

Fig 4(c) Recordings for the Mean Annual Temperature (°C) (MAT) for stations in Eastern Africa (1900-2022)

Fig4(d) Recordings for the Mean Annual Temperature (°C) (MAT) for stations in Eastern Africa (1950-2022) that had more than 20 annual recordings