@Article{MartinsBoDaDeDuJoRe:2010:EsAlDo,
               author = "Martins, Flora da Silva Ramos Vieira and Bock, J{\'e}r{\^o}me 
                         and Dambrine, Etienne and Dez, Gr{\'e}gory and Dupouey, Jean-Luc 
                         and Jolly, Anne and Renaud, Jean-Pierre",
          affiliation = "{Instituto Nacional de Pesquisas Espaciais (INPE)} and Office 
                         National des For{\^e}ts, R\&D Department Velaine-en-Haye, France 
                         and E-mail: jerome.bock@onf.fr Forest Ecosystems Biogeochemistry 
                         Unit, INRA, Champenoux, France and Office National des 
                         For{\^e}ts, R\&D Department Velaine-en-Haye, France and Forest 
                         Ecology \& Ecophysiology Unit, INRA, Champenoux, France. and 
                         Forest Ecology \& Ecophysiology Unit, INRA, Champenoux, France 
                         and Service r{\'e}gional de l’Arch{\'e}ologie de Lorraine, Metz, 
                         Franc",
                title = "Estimativa da altura dominante em povoamentos dec{\'{\i}}duos 
                         atrav{\'e}s de dados LIDAR com m{\'u}ltiplos retornos",
              journal = "Ambi{\^e}ncia",
                 year = "2010",
               volume = "6",
                pages = "115--126",
             keywords = "sensoriamento remoto, LIDAR, {\'{\i}}ndice de s{\'{\i}}tio, 
                         altura dominante, povoamentos dec{\'{\i}}duos.",
             abstract = "A altura dominante (Ho) de 120 parcelas georreferenciadas (600mē 
                         cada) foi estimada atrav{\'e}s de diversas medidas de 
                         distribui{\c{c}}{\~a}o, m{\'a}ximas locais e densidade derivada 
                         da varredura a laser com pequena impress{\~a}o obtida em um 
                         povoamento florestal dec{\'{\i}}duo irregular. As parcelas foram 
                         distribu{\'{\i}}das de forma a contemplar toda a variabilidade 
                         de alturas dominantes e o tipo de regenera{\c{c}}{\~a}o 
                         existente nos 112kmē da {\'a}rea estudada, localizada na floresta 
                         de Haye, Fran{\c{c}}a. O modelo constru{\'{\i}}do com 
                         vari{\'a}veis de m{\'a}xima local (Hmv5) e densidade de retornos 
                         (d9) foi capaz de estimar a Ho com alta acur{\'a}cia, al{\'e}m 
                         de ser independente do tipo de povoamento, o que possibilitou sua 
                         aplica{\c{c}}{\~a}o em toda a floresta. A valida{\c{c}}{\~a}o 
                         cruzada do modelo final mostrou que este explicou 98% da 
                         variabilidade observada nas parcelas em campo, com um RMSE de 
                         0.77m (3.31%). Nenhum efeito dr{\'a}stico da escolha do MNT, 
                         densidade de retornos ou do posicionamento das parcelas foi 
                         detectado no modelo, sugerindo grande estabilidade. Abstract The 
                         mean dominant heights of 120 georeferenced field sample plots 
                         (600mē each) were estimated from a range of canopy densities, 
                         distributions and local maxima metrics derived by a 
                         small-footprint laser scanner over various deciduous forest stands 
                         using regression analysis. The sample plots were distributed in 
                         order to better represent the variability in stand dominant height 
                         and regeneration practices throughout a 112kmē study areas in Haye 
                         forest, France. The model constructed with local maxima (Hmv5) and 
                         a LIDAR density metric (d9) was able to estimate Ho with a very 
                         high accuracy and was not sensitive to stand types. 
                         Cross-validation showed that the final model explained 98% of the 
                         variability in ground-truth dominant height, with a RMSE of 0.77m 
                         (3.31%). No drastic effects of DTM, echo densities, or positioning 
                         errors were found in the models.",
                 issn = "1808-0251",
                label = "lattes: 3335208724599495 1 MartinsBoDaDeDuJoRe:2010:EsAlDo",
             language = "en",
           targetfile = "888-4196-1-PB.pdf",
                  url = "http://revistas.unicentro.br/index.php/ambiencia/article/view/888",
        urlaccessdate = "21 maio 2024"
}