Decision tree classification of land cover from remotely sensed data

Decision tree classification algorithms have significant potential for land cover mapping problems and have not been tested in detail by the remote sensing community relative to more conventional pattern recognition techniques such as maximum likelihood classification. In this paper, we present several types of decision tree classification algorithms arid evaluate them on three different remote sensing data sets. The decision tree classification algorithms tested include an univariate decision tree, a multivariate decision tree, and a hybrid decision tree capable of including several different types of classification algorithms within a single decision tree structure. Classification accuracies produced by each of these decision tree algorithms are compared with both maximum likelihood and linear discriminant function classifiers. Results from this analysis show that the decision tree algorithms consistently outperform the maximum likelihood and linear discriminant function classifiers in regard to classf — cation accuracy. In particular, the hybrid tree consistently produced the highest classification accuracies for the data sets tested. More generally, the results from this work show that decision trees have several advantages for remote sensing applications by virtue of their relatively simple, explicit, and intuitive classification structure. Further, decision tree algorithms are strictly nonparametric and, therefore, make no assumptions regarding the distribution of input data, and are flexible and robust with respect to nonlinear and noisy relations among input features and class labels.

Changes in individual plant traits and biomass allocation in alpine meadow with elevation variation on the Qinghai-Tibetan Plateau

Plant traits and individual plant biomass allocation of 57 perennial herbaceous species, belonging to three common functional groups (forbs, grasses and sedges) at subalpine (3700 m ASL), alpine (4300 m ASL) and subnival (>= 5000 m ASL) sites were examined to test the hypothesis that at high altitudes, plants reduce the proportion of aboveground parts and allocate more biomass to belowground parts, especially storage organs, as altitude increases, so as to geminate and resist environmental stress. However, results indicate that some divergence in biomass allocation exists among organs. With increasing altitude, the mean fractions of total biomass allocated to aboveground parts decreased. The mean fractions of total biomass allocation to storage organs at the subalpine site (7%+/- 2% S.E.) were distinct from those at the alpine (23%+/- 6%) and subnival (21%+/- 6%) sites, while the proportions of green leaves at all altitudes remained almost constant. At 4300 m and 5000 m, the mean fractions of flower stems decreased by 45% and 41%, respectively, while fine roots increased by 86% and 102%, respectively. Specific leaf areas and leaf areas of forbs and grasses deceased with rising elevation, while sedges showed opposite trends. For all three functional groups, leaf area ratio and leaf area root mass ratio decreased, while fine root biomass increased at higher altitudes. Biomass allocation patterns of alpine plants were characterized by a reduction in aboveground reproductive organs and enlargement of fine roots, while the proportion of leaves remained stable. It was beneficial for high altitude plants to compensate carbon gain and nutrient uptake under low temperature and limited nutrients by stabilizing biomass investment to photosynthetic structures and increasing the absorption surface area of fine roots. In contrast to forbs and grasses that had high mycorrhizal infection, sedges had higher single leaf area and more root fraction, especially fine roots.