🌍 Bivariate Climate Map of Tanzania Using CHELSA Data in QGIS

This project demonstrates how to create a 2D bivariate map of Tanzania showing the spatial relationship between mean annual temperature and average monthly precipitation using CHELSA Bioclimatic data. The entire workflow is done using QGIS only — no coding or Python console is required.

---

🧰 Requirements

• QGIS 3.28 or later
• Enabled QGIS plugins:
  • ✅ Processing Toolbox
  • ✅ SAGA (for raster reclassification)
  • ✅ QuickOSM (optional, to obtain Tanzania boundary)

---

📦 Input Data

Dataset	Description
CHELSA_bio10_01.tif	Mean annual temperature (°C × 10)
CHELSA_bio10_12.tif	Total annual precipitation (mm)
Tanzania_boundary.shp	Tanzania national boundary (level 0)

🔗 Download CHELSA data from: https://chelsa-climate.org/downloads/

---

🔧 Workflow Steps

1. Clip CHELSA Rasters to Tanzania Boundary

Raster > Extraction > Clip Raster by Mask Layer

Repeat for both:

• CHELSA_bio10_01.tif → save as temp_clipped.tif
• CHELSA_bio10_12.tif → save as prec_clipped.tif

---

2. Convert Precipitation to Monthly Average

Raster > Raster Calculator

Expression:

"prec_clipped@1" / 30

Save as: prec_avg.tif

---

3. Reproject and Align Rasters (Optional but Recommended)

Processing > Toolbox > Warp (Reproject)

• Target CRS: EPSG:21037 (UTM Zone 37S - Tanzania)
• Resampling: Bilinear
• Align cell size and extent between rasters

---

4. Reclassify Both Rasters into 3 Quantile Classes

Step 1: Get Statistics

Raster > Miscellaneous > Raster Layer Statistics

Record min, max, and compute class breaks manually.

Step 2: Reclassify by Table (SAGA)

Processing > Toolbox > Reclassify by Table

• Use defined quantile breaks

• Assign:

  • Class 1 = Low
  • Class 2 = Medium
  • Class 3 = High

Save as:

• temp_class.tif
• prec_class.tif

---

5. Combine Reclassified Rasters into Bivariate Code

Raster > Raster Calculator

Expression:

"temp_class@1" * 10 + "prec_class@1"

Save as: bivariate.tif

This produces values from 11 to 33 representing each temperature-precipitation class combo.

---

6. Apply Bivariate Symbology

Layer Properties > Symbology

1. Change Render type: Singleband pseudocolor
2. Switch to Categorized
3. Click Classify to list values 11 to 33
4. Manually assign colors using the palette below

---

🎨 Bivariate Color Codes (DkBlue Palette)

Bivariate Value	Class Combo	Color Hex
11	Low Temp, Low Precipitation	#e8e8e8
12	Low Temp, Medium Precip	#dfb0d6
13	Low Temp, High Precip	#be64ac
21	Medium Temp, Low Precip	#ace4e4
22	Medium Temp, Medium Precip	#a5add3
23	Medium Temp, High Precip	#8c62aa
31	High Temp, Low Precip	#5ac8c8
32	High Temp, Medium Precip	#5698b9
33	High Temp, High Precip	#3b4994

Assign these manually in the categorized symbology panel in QGIS.

---

7. Add Map Layout and Export

Project > New Print Layout

• Add Map Frame

• Optional: Add custom bivariate legend (3×3 grid)

• Add:

  • Title: Tanzania: Temperature and Precipitation
  • Subtitle: CHELSA Bioclim (1981–2010), Monthly Avg. Precipitation
  • Caption: Source: CHELSA | Author: Your Name

• Export as .png, .svg, or .pdf

---

🧾 Suggested Folder Structure

tanzania_bivariate_map/
├── CHELSA_bio10_01.tif
├── CHELSA_bio10_12.tif
├── Tanzania_boundary.shp
├── temp_clipped.tif
├── prec_clipped.tif
├── prec_avg.tif
├── temp_class.tif
├── prec_class.tif
├── bivariate.tif
├── qgis_project.qgz
└── README.md

---

📚 References

• CHELSA Data: https://chelsa-climate.org
• QGIS: https://qgis.org
• Bivariate Colors: Inspired by biscale::bi_pal("DkBlue")
• Mapping Credits: Adapted by Milos Popovic (original R workflow)

---

🏁 Output

A clean 2D bivariate raster map showing the spatial co-distribution of temperature and precipitation across Tanzania.

---