Maria J. Diamantopoulou

Estimating tree bole volume using artificial neural network models for four species in Turkey

Tree bole volumes of 89 Scots pine (Pinus sylvestris L.), 96 Brutian pine (Pinus brutia Ten.), 107 Cilicica fir (Abies cilicica Carr.) and 67 Cedar of Lebanon (Cedrus libani A. Rich.) trees were estimated using Artificial Neural Network (ANN) models. Neural networks offer a number of advantages including the ability to implicitly detect complex nonlinear relationships between input and output variables, which is very helpful in tree volume modeling. Two different neural network architectures were used and produced the Back propagation (BPANN) and the Cascade Correlation (CCANN) Artificial Neural Network models. In addition, tree bole volume estimates were compared to other established tree bole volume estimation techniques including the centroid method, taper equations, and existing standard volume tables. An overview of the features of ANNs and traditional methods is presented and the advantages and limitations of each one of them are discussed. For validation purposes, actual volumes were determined by aggregating the volumes of measured short sections (average 1 meter) of the tree bole using Smalians formula. The results reported in this research suggest that the selected cascade correlation artificial neural network (CCANN) models are reliable for estimating the tree bole volume of the four examined tree species since they gave unbiased results and were superior to almost all methods in terms of error (%) expressed as the mean of the percentage errors.

The use of a Neural Network technique for the prediction of water quality parameters

This paper is concerned with the use of Neural Network models for the prediction of water quality parameters in rivers. The procedure that should be followed in the development of such models is outlined. Artificial Neural Networks (ANNs) were developed for the prediction of the monthly values of three water quality parameters of the Strymon river at a station located in Sidirokastro Bridge near the Greek — Bulgarian borders by using the monthly values of the other existing water quality parameters as input variables. The monthly data of thirteen parameters and the discharge, at the Sidirokastro station, for the time period 1980–1990 were selected for this analysis. The results demonstrate the ability of the appropriate ANN models for the prediction of water quality parameters. This provides a very useful tool for filling the missing values that is a very serious problem in most of the Greek monitoring stations.

Filling gaps in diameter measurements on standing tree boles in the urban forest of Thessaloniki, Greece

Missing data are omnipresent in forestry research, and this poses problems in the analysis of primary data. Many statistical problems have been viewed as missing data problems. To cope with incomplete data, several methods are currently being used. They are all based on assumptions some of which might not be valid in a particular case. The choice mostly depends on the objective of the study. Considerable mensuration research is motivated by the need for yield projections that can support forest management decisions. This paper is focused on a new approach for filling gaps in diameter measurements on standing tree boles. Dealing with this problem, an attempt was made to examine the applicability of artificial neural network models for missing data estimation and to use the estimated values in the subsequent analysis. The procedure that should be followed in the development of such models is outlined. The results show good performance of the examined ANN models compared to regression treatments for missing data and ANN models demonstrate their adequacy and potential for filling gaps in diameter measurements on standing tree boles. The ANN models applied in this study are sufficiently general and have great potential to be applicable for estimating the missing values of many variables in environmental applications.

Estimating breast height diameter and volume from stump diameter for three economically important species in Turkey.

Abstract Predicting volume directly from stump dimensions is useful in many situations, such as timber trespass. In the present study, equations for predicting diameter at breast height (dbh) and tree volume from stump diameter outside bark were developed for three economically important tree species in the forest region of Bucak, Turkey. Diameter at breast height was estimated with relatively high accuracy using a simple linear model. Tree volume was estimated with high precision using an exponential equation. Weighted linear and non-linear least squares methods were used to consider heteroscedasticity observed in the volume–stump diameter relationships. The results of a non-linear extra sum of squares method and of the F test indicated that different equations are needed for estimating dbh and tree volume from stump diameter outside bark for different species.

Tree-bark volume prediction via machine learning: A case study based on black alder’s tree-bark production

Abstract Tree-bark volume estimation is a multi-faceted problem and at the same time of vital importance in the area of forest resources management. This importance relies on the fact that it constitutes a key variable for accurately assessing timber quantities, while at the same time its use has been spread as a soil-covering product or as a soil fertilizer or as a substitutional medicinal product. Consequently, due to its substantial economic impact, the accurate prediction of the tree-bark volume is of utmost importance. In this study, we propose three bark volume prediction models for black alder trees (Alnus glutinosa (L.) Gaertn subsp. barbata (C.A. Mey.) Yalt.) each targeting a different creation source of the black alder forest. Hence, we used data from naturally regenerated, plantation and coppice stand types. 1334 stem analysis data were collected for three different stand types. Two different modeling techniques were used, the weighted nonlinear regression and the e-support vector regression techniques. These two modeling approaches were selected due to the fact that the need to handle regression analysis problems (noise in the data, high variability and/or non-normal distributions) is essential. The state-of-the-art approach suggests the usage of machine learning techniques in an effort to build reliable and robust models able to deal with complex environmental problems. An overall illustration of the precision obtained by the constructed models was conducted by statistical criteria such as the root mean square error, the correlation coefficient, the Furnival’s index of fit and the Akaike’s information criterion. Although the estimation and prediction errors of the two different modeling techniques seem to be close in pure numbers, the e-support vector regression models gave the most accurate results for all stand types as compared to the nonlinear regression. Based on the results obtained from this study, the constructed e-support vector regression models for modeling tree-bark volume showed a great ability to generalize, and thus worth considering as an alternative to regression modeling that enables increasing our ability for successful forest management.