Nur Algeet Abarquero
Research Subjects
Nur Algeet Abarquero has a total of 13 co-authors. Here are the top co-authors by number of shared publications:
- Miguel Marchamalo
Department of Land Morphology and Engineering, Universidad Politécnica de Madrid, 28040 Madrid, Spain
2 shared publications - Alfredo Fernández-Landa
Agresta Sociedad Cooperativa, 28012 Madrid, Spain
1 shared publication - Isabel González González
Agresta S. COOP., España
1 shared publication Javier Bonatti
1 shared publication- Jessica Esteban Cava
Agresta S. COOP., España
1 shared publication Jesús Fernández-Moya
1 shared publication
Latest Publications
- Open Access
- 18 March 2026
- Published by Universidad de Extremadura - Servicio de Publicaciones in Congresos UEX, actas de congresos
- No. 2
Clasificación automatizada basada en series temporales y variables estructurales a partir de sensores remotos (S2, S1 y LiDAR).
La clasificación precisa de las masas forestales es esencial para la gestión sostenible y la modelización de riesgos naturales. Este estudio evalúa la capacidad de discriminación de tres especies forestales (Pinus sylvestris, Quercus pyrenaica y Fagus sylvatica) en el Monte de Utilidad Pública de Villoslada de Cameros (La Rioja), mediante un enfoque de teledetección multitemporal y multisensor. Se integraron datos ópticos (Sentinel-2), de radar (Sentinel-1) y estructurales (LiDAR-PNOA) para caracterizar la fenología y estructura de la vegetación. La metodología partió de parcelas monoespecíficas del Inventario Forestal Nacional (IFN4) y datos locales, aplicando un sobremuestreo espacial mediante puntos aleatorios para corregir el desbalanceo de clases entre coníferas y frondosas. Tras generar un vector de 416 características, se utilizó el algoritmo de Eliminación Recursiva de Características con Validación Cruzada (RFECV) para seleccionar las 20 variables más robustas, optimizando el modelo. Los resultados mostraron una precisión global del 96,1%, identificando como predictores críticos la banda del Red-Edge (B6), el índice de Verdor del Tasseled Cap (TCG) y la polarización VH. Aunque la separación entre coníferas y frondosas fue excelente, se observó una ligera confusión entre las especies de frondosas debido a su similitud fenológica. Se concluye que las series temporales y las variables estructurales podrían ser determinante para mejorar la precisión en la automatización de la identificación de especies en inventarios forestales a gran escala.
- 18 March 2026
- Published by Universidad de Extremadura - Servicio de Publicaciones in Congresos UEX, actas de congresos
- No. 2
Open AccessClasificación automatizada basada en series temporales y variables estructurales a partir de sensores remotos (S2, S1 y LiDAR).
La clasificación precisa de las masas forestales es esencial para la gestión sostenible y la modelización de riesgos naturales. Este estudio evalúa la capacidad de discriminación de tres especies forestales (Pinus sylvestris, Quercus pyrenaica y Fagus sylvatica) en el Monte de Utilidad Pública de Villoslada de Cameros (La Rioja), mediante un enfoque de teledetección multitemporal y multisensor. Se integraron datos ópticos (Sentinel-2), de radar (Sentinel-1) y estructurales (LiDAR-PNOA) para caracterizar la fenología y estructura de la vegetación. La metodología partió de parcelas monoespecíficas del Inventario Forestal Nacional (IFN4) y datos locales, aplicando un sobremuestreo espacial mediante puntos aleatorios para corregir el desbalanceo de clases entre coníferas y frondosas. Tras generar un vector de 416 características, se utilizó el algoritmo de Eliminación Recursiva de Características con Validación Cruzada (RFECV) para seleccionar las 20 variables más robustas, optimizando el modelo. Los resultados mostraron una precisión global del 96,1%, identificando como predictores críticos la banda del Red-Edge (B6), el índice de Verdor del Tasseled Cap (TCG) y la polarización VH. Aunque la separación entre coníferas y frondosas fue excelente, se observó una ligera confusión entre las especies de frondosas debido a su similitud fenológica. Se concluye que las series temporales y las variables estructurales podrían ser determinante para mejorar la precisión en la automatización de la identificación de especies en inventarios forestales a gran escala.
- Open Access
- 28 February 2017
- Published by Italian Society of Sivilculture and Forest Ecology (SISEF) in iForest - Biogeosciences and Forestry
- Vol. 10 (1) , 259-266
High resolution biomass mapping in tropical forests with LiDAR-derived Digital Models: Poás Volcano National Park (Costa Rica)
Tropical forests play a key role in global carbon cycle. Reducing Emissions from Deforestation and forest Degradation (REDD+) program requires reliable mechanisms for Monitoring, Reporting and Verification (MRV). In this regard, new methods must be developed using updated technologies to assess carbon stocks. The combination of LiDAR technology and in situ forest networks allows the estimation of biomass with high resolution in low data environments, such as tropical countries. However, the evaluation of current LiDAR methods of biomass inventory, and the development of new methodologies to reduce uncertainty and increase accuracy, is still needed. Our aim is to evaluate new methodologies of spatially explicit LiDAR biomass inventories based on local and general plot-aggregate allometry. For this purpose, 25 field plots were inventoried, covering the structural and ecological variability of Poás Volcano National Park (Costa Rica). Important differences were detected in the estimation of aboveground biomass (92.74 t ha-1 considering the mean value of plot sample) depending on the chosen tree allometry. We validated the general aboveground biomass plot-aggregate allometry proposed by Asner & Mascaro (2014) in our study area, and we fitted two specific models for Poás forests. Both locals and general models depend on LiDAR top-of-canopy height (TCH), basal area (BA) and wood density. Small deviations in the wood density plot sample (0.60 ± 0.05) indicated that a single wood density constant value could be used throughout the study area. A BA-TCH origin forced linear model was fitted to estimate basal area, as suggested by the general methodology. Poás forest has a larger biomass density for the same THC compared to the rest of the forests previously studied, and shows that the BA-TCH relationship might have different trends in each life zone. Our results confirm that the general plot-aggregate methodology can be easily and reliably applied as aboveground biomass in a new area could be estimated by only measuring BA in field plots to obtain a local BA-TCH regression. For both local and general methods, the estimation of BA is critical. Therefore, the definition of precise basal area field measurement procedures is decisive to achieve reliable results in future studies.
- 28 February 2017
- Published by Italian Society of Sivilculture and Forest Ecology (SISEF) in iForest - Biogeosciences and Forestry
- Vol. 10 (1) , 259-266
Open AccessHigh resolution biomass mapping in tropical forests with LiDAR-derived Digital Models: Poás Volcano National Park (Costa Rica)
Tropical forests play a key role in global carbon cycle. Reducing Emissions from Deforestation and forest Degradation (REDD+) program requires reliable mechanisms for Monitoring, Reporting and Verification (MRV). In this regard, new methods must be developed using updated technologies to assess carbon stocks. The combination of LiDAR technology and in situ forest networks allows the estimation of biomass with high resolution in low data environments, such as tropical countries. However, the evaluation of current LiDAR methods of biomass inventory, and the development of new methodologies to reduce uncertainty and increase accuracy, is still needed. Our aim is to evaluate new methodologies of spatially explicit LiDAR biomass inventories based on local and general plot-aggregate allometry. For this purpose, 25 field plots were inventoried, covering the structural and ecological variability of Poás Volcano National Park (Costa Rica). Important differences were detected in the estimation of aboveground biomass (92.74 t ha-1 considering the mean value of plot sample) depending on the chosen tree allometry. We validated the general aboveground biomass plot-aggregate allometry proposed by Asner & Mascaro (2014) in our study area, and we fitted two specific models for Poás forests. Both locals and general models depend on LiDAR top-of-canopy height (TCH), basal area (BA) and wood density. Small deviations in the wood density plot sample (0.60 ± 0.05) indicated that a single wood density constant value could be used throughout the study area. A BA-TCH origin forced linear model was fitted to estimate basal area, as suggested by the general methodology. Poás forest has a larger biomass density for the same THC compared to the rest of the forests previously studied, and shows that the BA-TCH relationship might have different trends in each life zone. Our results confirm that the general plot-aggregate methodology can be easily and reliably applied as aboveground biomass in a new area could be estimated by only measuring BA in field plots to obtain a local BA-TCH regression. For both local and general methods, the estimation of BA is critical. Therefore, the definition of precise basal area field measurement procedures is decisive to achieve reliable results in future studies.