Current issue: 58(2)

Under compilation: 58(3)

Scopus CiteScore 2021: 2.8
Scopus ranking of open access forestry journals: 8^th
PlanS compliant

Articles containing the keyword 'k-NN'

Category : Research article

article id 10183, category Research article

Tomi Karjalainen, Petteri Packalen, Janne Räty, Matti Maltamo. (2019). Predicting factual sawlog volumes in Scots pine dominated forests using airborne laser scanning data. Silva Fennica vol. 53 no. 4 article id 10183. https://doi.org/10.14214/sf.10183

Keywords: Pinus sylvestris; area based approach; k-NN; sawlog

Highlights: We predicted visually bucked factual sawlog volumes at the 30 × 30 m plot-level with several alternatives; The lowest root mean squared error value of approximately 21% was obtained with a linear mixed-effects model that employed factual sawlog volume as a response variable and airborne laser scanning metrics as predictors; The sawlog reduction model commonly used in Finland performed poorly.

Abstract | Full text in HTML | Full text in PDF | Author Info

The aim in the study was to compare alternatives for the prediction of factual sawlog volumes using airborne laser scanning (ALS) data in Scots pine (Pinus sylvestris L.) dominated forests in eastern Finland. Accurate estimates of factual sawlog volume are desirable to ease the planning of harvesting operations. The factual sawlog volume of pines was derived from visual bucking, i.e. a procedure where the defects were located on each stem during sample plot measurements. For other species, the theoretical sawlog volume was considered also as the factual sawlog volume due to data restrictions. We predicted factual sawlog volume with eight alternatives that were based on either linear mixed-effects models or k-nearest neighbour imputations. An existing sawlog reduction model, commonly used in Finland, was also tested individually and combined with a number of the alternatives, and site type information was also utilised. Model fitting and prediction was implemented at the 15 × 15 m level, but accuracy was assessed at the 30 × 30 m level. The relative root mean squared error (RMSE%) values for the factual sawlog volume predictions varied between 20.9% and 33.5%, and the best accuracy was obtained with a linear mixed-effects model. These results indicate that factual sawlog volumes in Scots pine dominated forests can be predicted with reasonable accuracy with ALS data.

Karjalainen, University of Eastern Finland, School of Forest Sciences, P.O. Box 111, FI-80101 Joensuu, Finland E-mail: tomikar@uef.fi
Packalen, University of Eastern Finland, School of Forest Sciences, P.O. Box 111, FI-80101 Joensuu, Finland E-mail: petteri.packalen@uef.fi
Räty, University of Eastern Finland, School of Forest Sciences, P.O. Box 111, FI-80101 Joensuu, Finland E-mail: janne.raty@uef.fi
Maltamo, University of Eastern Finland, School of Forest Sciences, P.O. Box 111, FI-80101 Joensuu, Finland E-mail: matti.maltamo@uef.fi

article id 1568, category Research article

Jouni Siipilehto, Harri Lindeman, Mikko Vastaranta, Xiaowei Yu, Jori Uusitalo. (2016). Reliability of the predicted stand structure for clear-cut stands using optional methods: airborne laser scanning-based methods, smartphone-based forest inventory application Trestima and pre-harvest measurement tool EMO. Silva Fennica vol. 50 no. 3 article id 1568. https://doi.org/10.14214/sf.1568

Keywords: forest inventory; diameter distribution; Weibull; area-based approach; parameter recovery; k-NN estimation

Highlights: An airborne laser scanning grid-based approach for determining stand structure enabled bi- or multimodal predicted distributions that fitted well to the ground-truth harvester data; EMO and Trestima applications needed stand-specific inventory for sample measurements or sample photos, respectively, and at their best, provided superior accuracy for predicting certain stand characteristics.

Abstract | Full text in HTML | Full text in PDF | Author Info

Accurate timber assortment information is required before cuttings to optimize wood allocation and logging activities. Timber assortments can be derived from diameter-height distribution that is most often predicted from the stand characteristics provided by forest inventory. The aim of this study was to assess and compare the accuracy of three different pre-harvest inventory methods in predicting the structure of mainly Scots pine-dominated, clear-cut stands. The investigated methods were an area-based approach (ABA) based on airborne laser scanning data, the smartphone-based forest inventory Trestima app and the more conventional pre-harvest inventory method called EMO. The estimates of diameter-height distributions based on each method were compared to accurate tree taper data measured and registered by the harvester’s measurement systems during the final cut. According to our results, grid-level ABA and Trestima were generally the most accurate methods for predicting diameter-height distribution. ABA provides predictions for systematic 16 m × 16 m grids from which stand-wise characteristics are aggregated. In order to enable multimodal stand-wise distributions, distributions must be predicted for each grid cell and then aggregated for the stand level, instead of predicting a distribution from the aggregated stand-level characteristics. Trestima required a sufficient sample for reliable results. EMO provided accurate results for the dominating Scots pine but, it could not capture minor admixtures. ABA seemed rather trustworthy in predicting stand characteristics and diameter distribution of standing trees prior to harvesting. Therefore, if up-to-date ABA information is available, only limited benefits can be obtained from stand-specific inventory using Trestima or EMO in mature pine or spruce-dominated forests.

Siipilehto, Natural Research Institute Finland (Luke), Management and Production of Renewable Resources, P.O. Box 18, FI-01301 Vantaa, Finland E-mail: jouni.siipilehto@luke.fi
Lindeman, Natural Research Institute Finland, Green Technology, Kaironiementie 15, 39700 Parkano E-mail: harri.lindeman@luke.fi
Vastaranta, University of Helsinki, Department of Forest Sciences, P.O. Box 62 (Viikinkaari 11), FI-00014 University of Helsinki E-mail: mikko.vastaranta@helsinki.fi
Yu, Finnish Geospatial Research Institute (FGI), Department of Remote Sensing and Photogrammetry, National Land Survey of Finland, P.O. Box 15 (Geodeetinrinne 2), FI-02431, Masala, Finland E-mail: xiaowei.yu@maanmittauslaitos.fi
Uusitalo, Natural Research Institute Finland, Green Technology, Kaironiementie 15, 39700 Parkano E-mail: jori.uusitalo@luke.fi

article id 1218, category Research article

Mikko Niemi, Mikko Vastaranta, Jussi Peuhkurinen, Markus Holopainen. (2015). Forest inventory attribute prediction using airborne laser scanning in low-productive forestry-drained boreal peatlands. Silva Fennica vol. 49 no. 2 article id 1218. https://doi.org/10.14214/sf.1218

Keywords: remote sensing; forest technology; forest management planning; mapping; k-NN estimation; random forests

Highlights: Following current forest inventory practises, stem volume was predicted in low-productive drained peatlands (LPDPs) with a root mean square error (RMSE) of 13.7 m³ ha^–1; When 30 reference plots measured from LPDPs were added to the prediction, RMSE was decreased to 10.0 m³ ha^–1; Additional reference plots from LPDPs did not affect the forest inventory attribute predictions in productive forests.

Abstract | Full text in HTML | Full text in PDF | Author Info

Nearly 30% of Finland’s land area is covered by peatlands. In Northern parts of the country there is a significant amount of low-productive drained peatlands (LPDPs) where the average annual stem volume growth is less than 1 m³ ha^–1. The re-use of LPDPs has been considered thoroughly since Finnish forest legislation was updated and the forest regeneration prerequisite was removed from LPDPs in January 2014. Currently, forestry is one of the re-use alternatives, thus detailed forest resource information is required for allocating activities. However, current forest inventory practices have not been evaluated for sparse growing stocks (e.g., LPDPs). The purpose of our study was to evaluate the suitability of airborne laser scanning (ALS) for mapping forest inventory attributes in LPDPs. We used ALS data with a density of 0.8 pulses per m², 558 field-measured reference plots (500 from productive forests and 58 from LPDPs) and k nearest neighbour (k-NN) estimation. Our main aim was to study the sensitivity of predictions to the number of LPDP reference plots used in the k-NN estimation. When the reference data consisted of 500 plots from productive forest stands, the root mean square errors (RMSEs) for the prediction accuracy of Lorey’s height, basal area and stem volume were 1.4 m, 2.7 m² ha^–1 and 13.7 m³ ha^–1 in LPDPs, respectively. When 30 additional reference plots were allocated to LPDPs, the respective RMSEs were 1.1 m, 1.7 m² ha^–1 and 10.0 m³ ha^–1. Additional reference plot allocation did not affect the predictions in productive forest stands.

Niemi, Department of Forest Sciences, University of Helsinki, P.O. Box 27, FI-00014, Finland & Centre of Excellence in Laser Scanning Research, Finnish Geospatial Research Institute FGI, Geodeetinrinne 2, FI-02430, Finland E-mail: mikko.t.niemi@helsinki.fi
Vastaranta, Department of Forest Sciences, University of Helsinki, P.O. Box 27, FI-00014, Finland & Centre of Excellence in Laser Scanning Research, Finnish Geospatial Research Institute FGI, Geodeetinrinne 2, FI-02430, Finland E-mail: mikko.vastaranta@helsinki.fi
Peuhkurinen, Arbonaut Oy Ltd., Latokartanontie 7 A, FI-00700, Finland E-mail: jussi.peuhkurinen@arbonaut.com
Holopainen, Department of Forest Sciences, University of Helsinki, P.O. Box 27, FI-00014, Finland & Centre of Excellence in Laser Scanning Research, Finnish Geospatial Research Institute FGI, Geodeetinrinne 2, FI-02430, Finland E-mail: markus.holopainen@helsinki.fi

article id 975, category Research article

Renats Trubins, Ola Sallnäs. (2014). Categorical mapping from estimates of continuous forest attributes – classification and accuracy. Silva Fennica vol. 48 no. 2 article id 975. https://doi.org/10.14214/sf.975

Keywords: Sweden; land cover maps; forest type maps; map accuracy assessment; class membership probability; Bayesian network; k-NN estimates

Highlights: The paper presents an approach to classification and accuracy assessment of ad-hoc categorical maps based on existing spatial datasets with estimates of continuous forest variables; Pixel level class membership probabilities are estimated using a Bayesian network model.

Abstract | Full text in HTML | Full text in PDF | Author Info

Spatially explicit data on forest attributes is demanded for various research with landscape perspective. Existing datasets with estimates of continuous forest variables are often used as the basis for producing categorical forest type maps. Normally, this type of maps are used without knowing their accuracy. This paper presents a Bayesian network model for estimating pixel level class membership probabilities of thus derived categorical maps. Class membership probabilities can be used as a post-classification measure of map accuracy and in the process of map classification affecting the assignments of class labels. The method is applied in mapping deciduous dominated forests on the basis of the k-NN Sweden 2005 dataset in a study area in southern Sweden. The results indicate rather low accuracy for deciduous class regardless of the map classification method: 0.48 versus 0.50 in the maps classified without and with the use of the class membership probabilities given equal deciduous area. When probability-based classification is applied, the level of accuracy varies with the assumed map class proportions. Thus, when deciduous class area corresponding to the National Forest Inventory estimate was used, the accuracy of only 0.35 was obtained for the deciduous map class.

Trubins, Southern Swedish Forest Research Centre, Swedish University of Agricultural Sciences, Box 49, 230 53 Alnarp, Sweden E-mail: renats.trubins@slu.se
Sallnäs, Southern Swedish Forest Research Centre, Swedish University of Agricultural Sciences, Box 49, 230 53 Alnarp, Sweden E-mail: ola.sallnas@slu.se

Click this link to register to Silva Fennica.

If you are a registered user, log in to save your selected articles for later access.

Contents alert

Your selected articles

Follow @SilvaFennica