Selection of additional trainig data for improving accuracy of forest type classification using hyperspectral data

Abstract We investigated selection of stopover habitat by juvenile Swainson's Thrushes (Catharus ustulatus) during fall migration at a site along the northern California coast. The study site vegetation consisted mainly of coniferous forest (pine [Pinus] and spruce [Picea]), with interspersed patches of broadleaf forest (willow [Salix] and alder [Alnus]) in poorly drained swales. For 26 birds captured and radio-tracked in 2002 and 2003, the average minimum stopover duration was 8.9 ± 1.0 days. For 20 of these birds with a sufficient number of locations, the average home range size was 1.9 ± 0.3 ha. Thrushes showed no overall pattern of selection for forest type within the study area or for forest type used inside their home range. Fat and lean birds selected forest types similarly within the study area and their home ranges. However, locations occupied by lean birds had twice as much huckleberry (Vaccinium ovatum) shrub cover and were 1.3 times more concealed by vegetation than locations occupied by fat birds. There were 2.5 times more huckleberries at occupied than random locations, and locations occupied by lean birds had 2.1 times more berries overall than those frequented by fat birds. Fecal analyses confirmed that huckleberries were a commonly consumed food (70% of sampled thrushes), but also revealed that thrushes ate arthropods (87%) and wax myrtle (Myrica californica) bracteoles (43%). The overall lack of forest type selection coupled with differences between fat and lean birds in selection for cover and fruit abundance suggests that fat level may influence microsite selection.

Download Full-text

Land Cover and Forest Type Classification by Values of Vegetation Indices and Forest Structure of Tropical Lowland Forests in Central Vietnam

International Journal of Forestry Research ◽

10.1155/2020/8896310 ◽

2020 ◽

Vol 2020 ◽

pp. 1-18

Author(s):

Hung Nguyen Trong ◽

The Dung Nguyen ◽

Martin Kappas

Keyword(s):

Land Cover ◽

Vegetation Index ◽

Forest Type ◽

Vegetation Indices ◽

Landsat 8 ◽

Tropical Lowland ◽

Central Vietnam ◽

Lowland Forests ◽

Type Classification ◽

Sentinel 2

This paper aims to (i) optimize the application of multiple bands of satellite images for land cover classification by using random forest algorithms and (ii) assess correlations and regression of vegetation indices of a better-performed land cover classification image with vertical and horizontal structures of tropical lowland forests in Central Vietnam. In this study, we used Sentinel-2 and Landsat-8 to classify seven land cover classes of which three forest types were substratified as undisturbed, low disturbed, and disturbed forests where forest inventory of 90 plots, as ground-truth, was randomly sampled to measure forest tree parameters. A total of 3226 training points were sampled on seven land cover types. The performance of Landsat-8 showed out-of-bag error of 31.6%, overall accuracy of 68%, kappa of 67.5%, while Sentinel-2 showed out-of-bag error of 14.3% and overall accuracy of 85.7% and kappa of 83%. Ten vegetation indices of the better-performed image were extracted to find out (i) the correlation and regression of horizontal and vertical structures of trees and (ii) assess the variation values between ground-truthing plots and training sample plots in three forest types. The result of the t test on vegetation indices showed that six out of ten vegetation indices were significant at p<0.05. Seven vegetation indices had a correlation with the horizontal structure, but four vegetation indices, namely, Enhanced Vegetation Index, Perpendicular Vegetation Index, Difference Vegetation Index, and Transformed Normalized Difference Vegetation Index, had better correlations r = 0.66, 0.65, 0.65, 0.63 and regression results were of R2 = 0.44, 0.43, 0.43, and 0.40, respectively. The correlations of tree height were r = 0.46, 0.43, 0.43, and 0.49 and its regressions were of R2 = 0.21, 0.19, 0.18, and 0.24, respectively. The results show the possibility of using random forest algorithm with Sentinel-2 in forest type classification in line with vegetation indices application.

Download Full-text