Optimizing Field Data Collection for Individual Tree Attribute Predictions Using Active Learning Methods

Salim Malek,Damiano Gianelle,Terje Gobakken,Erik Næsset,Michele Dalponte,Franco Miglietta

doi:10.3390/rs11080949

Salim Malek, Damiano Gianelle + Show 4 more

Open Access

https://doi.org/10.3390/rs11080949

Copy DOI

Abstract

Light detection and ranging (lidar) data are nowadays a standard data source in studies related to forest ecology and environmental mapping. Medium/high point density lidar data allow to automatically detect individual tree crowns (ITCs), and they provide useful information to predict stem diameter and aboveground biomass of each tree represented by a detected ITC. However, acquisition of field data is necessary for the construction of prediction models that relate field data to lidar data and for validation of such models. When working at ITC level, field data collection is often expensive and time-consuming as accurate tree positions are needed. Active learning (AL) can be very useful in this context as it helps to select the optimal field trees to be measured, reducing the field data collection cost. In this study, we propose a new method of AL for regression based on the minimization of the field data collection cost in terms of distance to navigate between field sample trees, and accuracy in terms of root mean square error of the predictions. The developed method is applied to the prediction of diameter at breast heights (DBH) and aboveground biomass (AGB) of individual trees by using their height and crown diameter as independent variables and support vector regression. The proposed method was tested on two boreal forest datasets, and the obtained results show the effectiveness of the proposed selecting strategy to provide substantial improvements over the different iterations compared to a random selection. The obtained RMSE of DBH/AGB for the first dataset was 5.09 cm/95.5 kg with a cost equal to 8256/6173 m by using the proposed multi-objective method of selection. However, by using a random selection, the RMSE was 5.20 cm/102.1 kg with a cost equal to 28,391/30,086 m. The proposed approach can be efficient in order to get more accurate predictions with smaller costs, especially when a large forest area with no previous field data is subject to inventory and analysis.

Highlights

Having a precise description and representation of forest ecosystems in terms of carbon stock density and forest structure is an important key in international efforts to alleviate climate change [1,2]
Regarding experiments A and B, we can see that the random selection criteria provided the worst results, with a very high cost and a small improvement in terms of RMSE compared to the initial case
Using the MO Pareto optimization in order to find a compromise between the two criteria (Cost and Diversity), we can notice that the improvement in terms of RMSE is close to the one obtained by using only the Diversity criterion, but at a smaller cost

Summary

Introduction

Having a precise description and representation of forest ecosystems in terms of carbon stock density and forest structure is an important key in international efforts to alleviate climate change [1,2]. Numerous studies have been conducted in recent years on prediction modeling of such attributes using light detection and ranging (lidar) data [5,6] and adopting one of two main approaches, namely the area-based (ABA) [7,8] and individual tree crown (ITC) methods [9,10]. ITC approaches require field measurements of tree characteristics (e.g., species, DBH, height) and individual tree positions. An ideal situation would be to have an automatic method that, based on previous field data in areas with similar characteristics to the one subject to study, could help to reduce the inventory effort and cost by selecting the minimum number of trees needed in the subject area to construct a suitable prediction model

Objectives

Methods

Results

Discussion

Conclusion