Browsing by Author "Madundo, Sami Dawood"

Now showing 1 - 2 of 2
  • Thumbnail Image
    Item
    Comparison of multi-source remote sensing data for estimating and mapping above-ground biomass in the West Usambara tropical montane forests
    (Elsevier B.V., 2023-06) Madundo, Sami Dawood; Mauya, Ernest; Kilawe, Charles Joseph
    Above-ground biomass (AGB) estimation is important to better understand the carbon cy- cle and improve the efficiency of forest policy and management activities. AGB estimation models, using a combination of field data and remote sensing data, can largely replace traditional survey methods for measuring AGB. There are, however, critical steps for map- ping AGB based on satellite data with an acceptable degree of accuracy, such as choice of remote sensing data, the proper statistical modelling method, and remote sensing pre- dictor variables, at known field locations. This study sought to identify the optimal op- tical and synthetic aperture radar (SAR) remote sensing imagery from five sensors (Plan- etScope, Sentinel-2, Landsat 8 OLI, ALOS-2/PALSAR-2, and Sentinel-1) to model 159 field- based AGB values from two montane forests under semiparametric (Generalized Additive Model; GAM) and non-parametric (eXtreme Gradient Boosting; XGB) approaches using in- formation from four groups of predictor variables (spectral bands/polarizations, vegetation indices, textures, and a combination of all). The study’s results showed that PlanetScope (rRMSE = 69.19%; R 2 = 0.161) was the most precise optical sensor while ALOS-2/PALSAR-2 (rRMSE = 70.76; R 2 = 0.165) was the most precise amongst the SAR sensors. XGB mod- els generally resulted in those with lower prediction errors as compared to GAMs for the five sensors. Models having textures of vegetation indices and polarization bands achieved greater accuracy than models that incorporated spectral bands/polarizations and vegeta- tion indices only. The study recommends that PlanetScope and ALOS-2/PALSAR-2 remote sensing data using the XGB-based technique is an appropriate approach for accurate lo- cal and regional estimation of tropical forest AGB particularly for complex montane forest ecosystems.
  • Thumbnail Image
    Item
    Modelling and mapping forest above-ground biomass using earth observation data
    (Science Publishing Group, 2022) Madundo, Sami Dawood; Mauya, Ernest William; Lolila, Nandera Juma; Mchelu, Hadija Ahmed
    Accurate information on above-ground biomass (AGB) is important for sustainable forest management as well as for global initiatives aimed at combating climate change in the Tropics. In this study, AGB was estimated using a combination of field and Sentinel-2 earth observation data. The study was conducted at Magamba Nature Reserve in Lushoto district, Tanzania. Field plot-based AGB values were regressed against eighteen Sentinel-2 remote sensing variables (bands and vegetation indices) using Random Forest (RF) models based on centroid and weighted approaches. Results showed that the weighted model had the highest fit and precision (pseudo-R 2 = 0.21, rRMSE = 68.23%). A prediction map was produced with a mean AGB of 223.47 Mg ha -1 which was close to that of the field (225.19 Mg ha -1 ). Furthermore, the standard deviation of the AGB obtained from the map (i.e 174.04 Mg ha -1 ) was relatively lower as compared to the one obtained from the field-based measurements (i.e 97.42 Mg ha -1 ). This study demonstrated that Sentinel-2 imagery and RF-based regression techniques have potential to effectively support large scale estimation of forest AGB in the tropical rainforests.