Forest structure modeling with combined airborne hyperspectral and LiDAR data

Latifi, Hooman; Fassnacht, Fabian Ewald; Koch, Bárbara

doi:10.1016/j.rse.2012.01.015

Cited by 101 publications

(69 citation statements)

References 48 publications

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…and when it is suspected that parametric conditions are violated (e.g., reduced datasets). This finding agrees with other important results in the area (Latifi et al, 2012) and with previous work criticizing stepwiserelated selection techniques (Whittingham et al, 2006).…”

Section: Figsupporting

confidence: 93%

“…Thus, Latifi et al (2010) showed that the use of this type of algorithm may improve the results of non-parametric prediction techniques for stand timber volume and biomass. In the same line, Latifi et al (2012) showed that GAs may also play an important role in the selection of predictors not only in LiDAR but also fused with hyperspectral data. Li et al (2013) also proved GAs power when they are combined with a linear discriminant analysis to classify tree species.…”

Section: Introductionmentioning

confidence: 92%

“…In the case of the best-subset regression, a previous manifold selection is usually needed since computers may not work in the proper manner on sets with many predictors; for example, a 1-GHz processor could need about 80 h to deal with the variations of the 48 predictors used in this study (2 48 variations), but this number can easily be higher in data fusion (Latifi et al, 2012). Although human guidance could have caused the differences in the results, the problems related to the misuse of the F-test to assure models with statistically significant variables can again be the responsible for their limited performance since the F-test is used to generate models in which all of the variables are statistically different.…”

Section: Figmentioning

confidence: 99%

“…In the context of LiDAR regression, a GA has been defined as a variable-screening procedure, based on the principle of evolution by natural selection, which develops sets of variables that have evolved to fit a certain fitness function via cycles of a differential replication, recombination, and mutation procedure (Latifi et al, 2012). Thus, Latifi et al (2010) showed that the use of this type of algorithm may improve the results of non-parametric prediction techniques for stand timber volume and biomass.…”

Section: Introductionmentioning

confidence: 99%

See 3 more Smart Citations

Evolutionary feature selection to estimate forest stand variables using LiDAR

García-Gutiérrez

González-Ferreiro

Santos

et al. 2014

International Journal of Applied Earth Observation and Geoinfor

View full text Add to dashboard Cite

a b s t r a c tLight detection and ranging (LiDAR) has become an important tool in forestry. LiDAR-derived models are mostly developed by means of multiple linear regression (MLR) after stepwise selection of predictors. An increasing interest in machine learning and evolutionary computation has recently arisen to improve regression use in LiDAR data processing. Although evolutionary machine learning has already proven to be suitable for regression, evolutionary computation may also be applied to improve parametric models such as MLR. This paper provides a hybrid approach based on joint use of MLR and a novel genetic algorithm for the estimation of the main forest stand variables. We show a comparison between our genetic approach and other common methods of selecting predictors. The results obtained from several LiDAR datasets with different pulse densities in two areas of the Iberian Peninsula indicate that genetic algorithms perform better than the other methods statistically. Preliminary studies suggest that a lack of parametric conditions in field data and possible misuse of parametric tests may be the main reasons for the better performance of the genetic algorithm. This research confirms the findings of previous studies that outline the importance of evolutionary computation in the context of LiDAR analisys of forest data, especially when the size of fieldwork datatasets is reduced.

show abstract

Section: Figsupporting

confidence: 93%

Section: Introductionmentioning

confidence: 92%

Section: Figmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

See 2 more Smart Citations

Evolutionary feature selection to estimate forest stand variables using LiDAR

García-Gutiérrez

González-Ferreiro

Santos

et al. 2014

International Journal of Applied Earth Observation and Geoinfor

View full text Add to dashboard Cite

show abstract

“…A plausible approach to investigating forest belowground biomass at large scale is to establish a relationship between a number of dendrometric parameters related to the aboveground vegetation (e.g., tree diameter and height) and the belowground component of the total biomass (Drexhage and Colin 2001;Peichl and Arain 2007;Laganière et al 2010). Dendrometric parameters can be derived from varied sources of information ranging from field measurements to remotely sensed data and generally involve the use of allometric equations (Latifi et al 2012;Zolkos et al 2013;Kankare et al 2013;Ahmed et al 2013). Furthermore, few investigations have explored belowground biomass in the Mediterranean forest ecosystems with a specific focus on sclerophyllous vegetation (i.e., maquis and garrigue), despite the wide geographic distribution of this component (RuizPeinado et al 2013).…”

Section: Handling Editor: Shuqing Zhaomentioning

confidence: 99%

Estimating belowground biomass and root/shoot ratio of Phillyrea latifolia L. in the Mediterranean forest landscapes

Marziliano

Lafortezza

Medicamento

et al. 2015

Annals of Forest Science

View full text Add to dashboard Cite

& Key message This study developed and tested models to predict the belowground biomass and root/shoot ratio using aboveground field measures. The predictive power of such indirect measurement is useful for a rapid and reliable assessment of the biomass of the Mediterranean species. & Context Forest biomass estimation has been simplified by the availability of indirect methodologies for automatic measuring. However, most of the information on forest root systems is largely unexplored due to the difficulty in estimating belowground biomass (BGB) at large scale. A plausible approach to investigating forest BGB is to establish a relationship between a number of dendrometric parameters related to the aboveground vegetation (AGB) (e.g., tree diameter and height) and the belowground component of the total biomass. & Aims This work presents findings for indirect measurements of BGB in the typical Mediterranean landscapes focusing on sclerophyllous vegetation, specifically Phillyrea latifolia L. The purpose of the present study is twofold: (a) to develop a model explaining the BGB distribution of P. latifolia based on field data for dendrometric parameters, and (b) to understand how the ratio between BGB and AGB varies according to stem diameter as a proxy of plant growth. & Methods A total of 50 P. latifolia plants were randomly selected in the study areas and considered for excavation. Individual plants were analyzed to determine AGB and BGB development. A number of models were developed and tested to predict the BGB and root/shoot ratio using aboveground field measures. Allometric equations were employed to predict the AGB and BGB and relative partitioning in the Phillyrea community. & Results Models for P. latifolia AGB and BGB estimation that include crown diameter and stem height measures augment the models' predictive power. When used alone, the predictive power of the root collar diameter appears to be overestimated, while its effect is stronger for a subset of observations with larger crown diameter and stem height. The root/shoot ratio values of plant species typically related to the Mediterranean context seem to be largely superior to the ratio values of trees and forests. & Conclusions The model is ideally suited to incorporate indirectly measured tree height for a rapid and reliable assessment of the biomass of single Mediterranean species. Further research might include replication of the same studies in different geographic areas of the Mediterranean and in-depth analyses of AGB.

show abstract

Mapping forest vegetation for the western United States using modified random forests imputation ofFIAforest plots

2016

View full text Add to dashboard Cite

Abstract. Maps of the number, size, and species of trees in forests across the western United States are desirable for many applications such as estimating terrestrial carbon resources, predicting tree mortality following wildfires, and for forest inventory. However, detailed mapping of trees for large areas is not feasible with current technologies, but statistical methods for matching the forest plot data with biophysical characteristics of the landscape offer a practical means to populate landscapes with a limited set of forest plot inventory data. We used a modified random forests approach with Landscape Fire and Resource Management Planning Tools (LANDFIRE) vegetation and biophysical predictors to impute plot data collected by the US Forest Service's Forest Inventory Analysis (FIA) to the landscape at 30-m grid resolution. This method imputes the plot with the best statistical match, according to a "forest" of decision trees, to each pixel of gridded landscape data. In this work, we used the LANDFIRE data set for gridded input because it is publicly available, offers seamless coverage of variables needed for fire models, and is consistent with other data sets, including burn probabilities and flame length probabilities generated for the continental United States. The main output of this project is a map of imputed plot identifiers at 30 × 30 m spatial resolution for the western United States that can be linked to the FIA databases to produce tree-level maps or to map other plot attributes. In addition, we used the imputed inventory data to generate maps of forest cover, forest height, and vegetation group at 30 × 30 m resolution for all forested pixels in the western United States, as a means of assessing the accuracy of our methodology. The results showed good correspondence between the target LANDFIRE data and the imputed plot data, with an overall within-class agreement of 79% for forest cover, 96% for forest height, and 92% for vegetation group.

show abstract

Forest structure modeling with combined airborne hyperspectral and LiDAR data

Cited by 101 publications

References 48 publications

Evolutionary feature selection to estimate forest stand variables using LiDAR

Evolutionary feature selection to estimate forest stand variables using LiDAR

Estimating belowground biomass and root/shoot ratio of Phillyrea latifolia L. in the Mediterranean forest landscapes

Mapping forest vegetation for the western United States using modified random forests imputation ofFIAforest plots

Contact Info

Product

Resources

About