Publicação
The use of remotely sensed data for forest biomass monitoring : a case of forest sites in north-eastern Armenia
| dc.contributor.advisor | Silva, Joel Dinis Baptista Ferreira da | |
| dc.contributor.advisor | Ghazaryan, Gohar | |
| dc.contributor.advisor | Guerrero, Ignacio | |
| dc.contributor.author | Khudinyan, Manvel | |
| dc.date.accessioned | 2019-03-18T19:24:11Z | |
| dc.date.available | 2019-03-18T19:24:11Z | |
| dc.date.issued | 2019-02-28 | |
| dc.description | Dissertation submitted in partial fulfilment of the requirements for the Degree of Master of Science in Geospatial Technologies | pt_PT |
| dc.description.abstract | In recent years there has been an increasing interest in the use of synthetic aperture radar (SAR) data and geospatial technologies for environmental monitoring․ Particularly, forest biomass evaluation was of high importance, as forests have a crucial role in global carbon emission. Within this study we evaluate the use of Sentinel 1 C-band multitemporal SAR data with combination of Alos Palsar L-band SAR and Sentinel 2 multispectral remote sensing (RS) data for mapping forest aboveground biomass (AGB) of dry subtropical forests in mountainous areas. Field observation from National Forest Inventory was used as a ground truth data. As the SAR data suffers greatly by the complex topography, a simple approach of aspect and slope information as forestry ancillary data was implemented directly in the regression model for the first time to mitigate the topography effect on radar backscattering value․ Dense time-series analysis allowed us to overcome the SAR saturation by the forest phenology and select the optimal C-band scene. Image texture measures of SAR data has been strongly related to the biomass distribution and has robustly contributed to the prediction․ Multilinear Stepwise Regression allowed to select and evaluate the most relevant variables for AGB. The prediction model combining RS with ancillary data explained the 62 % of variance with root-mean-square error of 56.6 t ha¯¹. The study also reveals that C-band SAR data on forest biomass prediction is limited due to their short wavelength. Further, the mountainous condition is a major constraint for AGB estimation. Additionally, this research demonstrates a positive outcome in forest AGB prediction with freely accessible RS data. | pt_PT |
| dc.identifier.tid | 202196305 | pt_PT |
| dc.identifier.uri | http://hdl.handle.net/10362/63694 | |
| dc.language.iso | eng | pt_PT |
| dc.rights.uri | http://creativecommons.org/licenses/by/4.0/ | pt_PT |
| dc.subject | Synthetic Aperture Radar | pt_PT |
| dc.subject | Forest | pt_PT |
| dc.subject | Aboveground Forest Biomass | pt_PT |
| dc.subject | Multilinear Stepwise regression | pt_PT |
| dc.subject | Sentinel 1 | pt_PT |
| dc.subject | Sentinel 2 | pt_PT |
| dc.subject | ALOS PALSAR | pt_PT |
| dc.subject | GLCM | pt_PT |
| dc.title | The use of remotely sensed data for forest biomass monitoring : a case of forest sites in north-eastern Armenia | pt_PT |
| dc.type | master thesis | |
| dspace.entity.type | Publication | |
| rcaap.rights | openAccess | pt_PT |
| rcaap.type | masterThesis | pt_PT |
| thesis.degree.name | Mestrado em Tecnologias Geoespaciais | pt_PT |