University of Twente Student Theses
Upscaling the assessment of biophysical forest parameters from high to medium resolution satellite imagery in Kazimzumbwi Forest Reserve, Coastal Region, Tanzania
Mushi, Seleboni John (2009) Upscaling the assessment of biophysical forest parameters from high to medium resolution satellite imagery in Kazimzumbwi Forest Reserve, Coastal Region, Tanzania.
|
PDF
2MB |
| Abstract: | Forests actively contribute to the world’s environmental stability and are used as economic resources to produce subsistence and industrial forest products. Remote sensing has been considered a low cost approach suited to broad scale forest inventory. Estimation of forest stand parameters such as biomass, stocking density, average crown diameter and species by group in a large area using remotely sensed data has considerable contribution to sustainable management and utility of natural resources. Biophysical forest properties have most commonly been estimated from remotely sensed data through use of vegetation indices. The objective of the study was to find the possibility of upscaling high to medium resolution imagery using Quickbird and ASTER images respectively. Ground field inventory revealed above ground biomass of 15.8 tonnes.ha-1, 4135 stems.ha-1, and basal area of 6.49m2.ha-1. Afzelia quanzensis was found to dominate the forest with 1186 stems.ha-1(28.7%). Crown diameter from multi resolution segmentation performed in high resolution imagery found relationships between dbh and biomass measured in the field for all trees in general, Albizia petersiana, Afzelia quanzensis and Brachystegia boehmii. It was not possible to perform upscaling from high to medium resolution imagery using spectral characteristics. Results from the relationships between Normalized Difference Vegetation Index (NDVI) from Quickbird image to Aster image were poor; R2 =0.1. Enhanced Vegetation Index (EVI) extracted from both images was also poor; R2 = 0.14. Differences in cell size between high resolution imagery (Quickbird) and medium resolution imagery (ASTER) contributed to failure. The Quickbird cell size does not present full object while the ASTER present full object(s). Key words: upscaling, multi-resolution segmentation, crown diameter, biophysical forest parameters, medium resolution imagery, miombo woodlands |
| Item Type: | Essay (Master) |
| Faculty: | ITC: Faculty of Geo-information Science and Earth Observation |
| Programme: | Geoinformation Science and Earth Observation MSc (75014) |
| Link to this item: | https://purl.utwente.nl/essays/93060 |
| Export this item as: | BibTeX EndNote HTML Citation Reference Manager |
Show download statistics for this publication
Show download statistics for this publicationRepository Staff Only: item control page