Application of shadow fraction models for estimating attributes of northern boreal forests

A shadow fraction method was developed previously for mapping forest attributes of northern black spruce (Picea mariana (Mill.) Britton, Sterns & Poggenb.) forests. This paper evaluates application of the method for (i) balsam fir stands (Abies balsamea (L.) Mill.), (ii) stands with higher volume and biomass than those of previous studies, and (iii) stands with a higher composition of deciduous trees and steeper slopes. Models developed for new test sites in (i) central Labrador and (ii) western Newfoundland were not statistically different from previous models for biomass, volume, and basal area. Rela- tive root mean square errors (RMSEr) for central Labrador were slightly lower than those found in other test sites (RMSEr: 24%-29%) but higher for western Newfoundland (RMSEr = 37%-43%), attributed to the higher upper limit of measured at- tributes and increased presence of deciduous trees. Results suggest that reasonable estimates can be generated for conifer forests of northeastern Canada; however, an alternative solution is needed where mixed and deciduous stands are prevalent. Measurement of ground plots over a wider range of species composition and forest structure is recommended for broader ap- plication to northern boreal forests and to further assess the potential role of the shadow fraction method in national-scale inventory programs. Resume : Une methode basee sur la fraction d'ombre a ete developpee precedemment pour cartographier les attributs des peuplements nordiques d'epinette noire (Picea mariana (Mill.) Britton, Sterns & Poggenb.). Cet article evalue l'application de la methode pour (i) les peuplements de sapin baumier (Abies balsamea (L.) Mill.), (ii) les peuplements dont le volume et la biomasse sont plus eleves que ceux des etudes precedentes et (iii) les peuplements situes sur des pentes plus abruptes et dont la proportion d'essences feuillues est plus elevee. Les modeles developpes pour les nouveaux sites d'essai dans (i )l e centre du Labrador et (ii )l 'ouest de Terre-Neuve n'etaient pas statistiquement differents des modeles precedents dans le cas de la biomasse, du volume et de la surface terriere. Les erreurs quadratiques moyennes relatives pour le centre du Labrador etaient legerement inferieures a celles qui ont ete calculees dans d'autres sites d'essai (24%-29%), mais elles etaient plus elevees pour l'ouest de Terre-Neuve (37%-43%). Cela serait du a la limite superieure plus elevee des attributs qui ont ete mesures et a la presence accrue de feuillus. Les resultats indiquent que des estimations raisonnables peuvent etre generees pour les forets de coniferes du nord-est du Canada, mais une solution alternative est necessaire la ou les peuplements mixtes et feuillus sont repandus. La mesure de placettes terrestres representant une large gamme de composition d'especes et de structure forestiere est recommandee pour permettre une application plus large aux forets boreales nordiques et continuer a evaluer le role potentiel de la methode proposee dans les programmes nationaux d'inventaire. (Traduit par la Redaction)

[1]  R. Houghton,et al.  Aboveground Forest Biomass and the Global Carbon Balance , 2005 .

[2]  F. Raulier,et al.  Canadian national tree aboveground biomass equations , 2005 .

[3]  R. McRoberts Using satellite imagery and the k-nearest neighbors technique as a bridge between strategic and management forest inventories , 2008 .

[4]  A. Beaudoin,et al.  A shadow fraction method for mapping biomass of northern boreal black spruce forests using QuickBird imagery , 2007 .

[5]  J. Susan Milton,et al.  Introduction to Probability and Statistics: Principles and Applications for Engineering and the Computing Sciences , 1990 .

[6]  W. Kurz,et al.  National level forest monitoring and modeling in Canada , 2004 .

[7]  C. Peng,et al.  Developing and Validating Nonlinear Height-Diameter Models for Major Tree Species of Ontario's Boreal Forests , 2001 .

[8]  I. Ozdemir,et al.  Estimating stem volume by tree crown area and tree shadow area extracted from pan‐sharpened Quickbird imagery in open Crimean juniper forests , 2008 .

[9]  Richard A. Fournier,et al.  Mapping aboveground tree biomass at the stand level from inventory information: test cases in Newfoundland and Quebec , 2003 .

[10]  J. Heiskanen,et al.  Biomass estimation over a large area based on standwise forest inventory data and ASTER and MODIS satellite data: A possibility to verify carbon inventories , 2007 .

[11]  D. Ruppert,et al.  Transformation and Weighting in Regression , 1988 .

[12]  Michael A. Wulder,et al.  Radiometric Image Processing , 2003 .

[13]  A. Beaudoin,et al.  Forest attribute estimation of northeastern Canadian forests using QuickBird imagery and a shadow fraction method , 2012 .

[14]  H. Burkhart,et al.  Top height definition and its effect on site index determination in thinned and unthinned loblolly pine plantations , 2002 .

[15]  D. Leckie,et al.  Forest inventory in Canada with emphasis on map production , 1995 .