Informations et ressources scientifiques
sur le développement des zones arides et semi-arides

Accueil du site → Master → Pays Bas → Aboveground woody biomass assessment in Serowe woodlands, Botswana.

International Institute for Geo-Information Science and Earth Observation (ITC) 2006

Aboveground woody biomass assessment in Serowe woodlands, Botswana.

Mabowe, B.R. (Boiki Raditsebe)

Titre : Aboveground woody biomass assessment in Serowe woodlands, Botswana.

Auteur : Mabowe, B.R. (Boiki Raditsebe)

Etablissement de soutenance : University of Twente International Institute for Geo-Information Science and Earth Observation (ITC)

Grade : Master of Science in Geo-Information Science and Earth Observation 2006

Quantitative assessment of aboveground woody biomass in terms of volume, fresh and or dry weight per unit area is a useful way to provide estimates of various tree components that can be harvested. This is more relevant to Botswana forests and woodlands since 53 % of the population depends on wood resources for energy purposes in the form of fuelwood. However, the approaches used to quantify the resource can be laborious and time consuming hence costly. It is because of this reason that data on aboveground woody biomass in Botswana is unreliable and scanty. Nonetheless, the technique of remote sensing has become common in forest investigation providing the only realistic and cost effective way of acquiring data over larger areas. The field data is used for validating the remote sensing data. However, one major problem is with the field part which is addressed prominently in this thesis. Two biomass estimation methods ; model stem method (Adhikari, 2005 ; Montes et al., 2000) and the sub-sampling method (de Gier, 1989) were compared in terms of time use efficiency and reliability. It was found that model stem method was very efficient in terms of time usage but not reliable. The sub-sampling method was also efficient in terms of time usage in the field, but was also found to be reliable and a better method in many ways compared to model stem method. Sub-sampling method provides in-the-field estimates of volume and fresh weight. Dry weight can be calculated after drying the same wood samples (disks) used to calculate the fresh woody biomass. A few light weight pieces of equipment are required for the application of this method and field work can be carried out by only two people. Tree species sampled were Terminalia sericea, Dichrostachys cinerea, Ochna pulchra, Burkea africana, Lonchocarpus nelsii, Boscia albitrunca and Acacia fleckii. During the development of biomass equations, it was found that there was no rationale for the development of species-specific equations. Hence mixed-species equation was developed using weighted curvilinear regression technique. This was facilitated by the use of POLYREG and REGDAT programs, resulting in a polynomial biomass equation based on the diameter at ankle height ( dah ). The equation was used to estimate aboveground woody biomass per plot (kg/m 2 ), using the inventory data collected in the field for 169 plots. Next, the potential of satellite based remote sensing techniques combined with biomass estimation method for quantitative assessment of aboveground woody biomass in the study area (Serowe woodlands of Botswana) was investigated. An IKONOS image of February, 2002 was used to derive spectral vegetation indices namely ; the Normalized Difference Vegetation Index (NDVI), Soil Adjusted Vegetation Index (SAVI), Enhanced Vegetation Index (EVI) and Perpendicular Vegetation Index (PVI). Vegetation maps were overlaid with the plot shapefile and the mean vegetation indices values were extracted and correlated with aboveground biomass (kg/m 2 ) per plot. However, the relationship between estimated aboveground woody biomass and vegetation indices derived from the IKONOS image showed poor correlation coefficients (R) of NDVI = 0.024, EVI = 0.083, SAVI = 0.077 , and PVI = 0.060. Therefore, estimating woody biomass using spectral data proved to be difficult. Hence the biomass map of the study area could not be produced. Though the potential of ground based biomass estimation combined with satellite based remote sensing proved difficult, the practical use of this thesis work can be found in providing adequate and reliable information regarding the woody biomass resource particularly fuelwood.

Version intégrale (ITC)

Page publiée le 5 février 2018