Estimating Tree Crown Area and Aboveground Biomass in Miombo Woodlands From High-Resolution RGB-Only Imagery

H. T. Mareya, Paradzayi Tagwireyi (Corresponding Author), H. Ndaimani, T. W. Gara, D. Gwenzi

Research output: Contribution to journalArticleAcademicpeer-review

8 Citations (Scopus)


Quantification of tree canopy area and aboveground biomass is essential for monitoring ecosystems' ecological functionalities, e.g., carbon sequestration and habitat provision. Miombo woodlands are vastly existent in developing countries that often lack resources to acquire LiDAR data or high spatiospectral resolution remote sensing data that have been proven to accurately estimate these structural attributes. This study explored the utility of freely available (via Google Maps) high (1-m) resolution red, green, and blue (RGB) satellite imagery in combination with object-based image analysis (OBIA) for estimating tree canopy area and aboveground biomass in Miombo woodlands. We randomly established 41 225-m 2 plots in Mukuvisi Woodland, Zimbabwe, and used RGB data with OBIA to estimate tree canopy area in those plots. We also field measured the height, canopy area, and trunk diameter at breast height of all trees that fell in those plots, then used the field data and a published allometric equation to estimate aboveground tree biomass (AGB). OBIA classification accuracy was high (Jaccard similarity index = 0.96). Data analysis showed significant positive linear relationship between AGB and field-measured canopy area (R 2 = 0.87, p <; 0.003), and significant exponential relationships between: 1) OBIA-derived canopy area and AGB (R 2 = 0.56, p <; 0.0001); and 2) field-measured canopy area and OBIA-derived canopy area (R 2 = 0.63, p <; 0.0001), and no significant differences (t = 19.67, df = 78, p = 0.28) between field-measured canopy area (×̅ = 187.11 m 2 , σ = 127.03) and OBIA-derived canopy area (×̅ = 163.00 m 2 , σ = 50.08). We conclude that RGB data with OBIA are suitable for estimating tree canopy area in Miombo woodlands for various low-accuracy purposes (e.g., biomass estimation).
Original languageEnglish
Pages (from-to)868-875
Number of pages8
JournalIEEE Journal of selected topics in applied earth observations and remote sensing
Issue number3
Publication statusPublished - 1 Mar 2018


  • ecology
  • forestry
  • geophysical image processing
  • image classification
  • remote sensing
  • vegetation
  • Google Maps
  • Jaccard similarity index
  • LiDAR data
  • Miombo woodlands
  • OBIA
  • OBIA classification accuracy
  • OBIA-derived canopy area
  • RGB data
  • Zimbabwe
  • aboveground tree biomass estimation
  • high resolution RGB satellite imagery
  • high resolution red green and blue satellite imagery
  • high spatiospectral resolution remote sensing data
  • high-resolution RGB-only imagery
  • object-based image analysis
  • tree canopy area
  • tree crown area
  • Biomass
  • Google
  • Monitoring
  • Remote sensing
  • Spatial resolution
  • Vegetation
  • And blue (RGB) imagery
  • Miombo
  • forest biomass
  • forest carbon mapping
  • green
  • object-based image analysis (OBIA)
  • red


Dive into the research topics of 'Estimating Tree Crown Area and Aboveground Biomass in Miombo Woodlands From High-Resolution RGB-Only Imagery'. Together they form a unique fingerprint.

Cite this