Search

Your search keyword '"Varvia, Petri"' showing total 15 results
Start Over Author "Varvia, Petri" Search Limiters Peer Reviewed
15 results on '"Varvia, Petri"'

1. Estimation of boreal forest biomass from ICESat-2 data using hierarchical hybrid inference

Author

Varvia, Petri, Saarela, Svetlana, Maltamo, Matti, Packalen, Petteri, Gobakken, Terje, Næsset, Erik, Ståhl, Göran, and Korhonen, Lauri

Subjects
Statistics - Applications
Abstract

The ICESat-2, launched in 2018, carries the ATLAS instrument, which is a photon-counting spaceborne lidar that provides strip samples over the terrain. While primarily designed for snow and ice monitoring, there has been a great interest in using ICESat-2 to predict forest above-ground biomass density (AGBD). As ICESat-2 is on a polar orbit, it provides good spatial coverage of boreal forests. The aim of this study is to evaluate the estimation of mean AGBD from ICESat-2 data using a hierarchical modeling approach combined with rigorous statistical inference. We propose a hierarchical hybrid inference approach for uncertainty quantification of the AGBD estimated from ICESat-2 lidar strips. Our approach models the errors coming from the multiple modeling steps, including the allometric models used for predicting tree-level AGB. For testing the procedure, we have data from two adjacent study sites, denoted Valtimo and Nurmes, of which Valtimo site is used for model training and Nurmes for validation. The ICESat-2 estimated mean AGBD in the Nurmes validation area was 63.2$\pm$1.9 Mg/ha (relative standard error of 2.9%). The local reference hierarchical model-based estimate obtained from wall-to-wall airborne lidar data was 63.9$\pm$0.6 Mg/ha (relative standard error of 1.0%). The reference estimate was within the 95% confidence interval of the ICESat-2 hierarchical hybrid estimate. The small standard errors indicate that the proposed method is useful for AGBD assessment. However, some sources of error were not accounted for in the study and thus the real uncertainties are probably slightly larger than those reported.

Published
2023

3. Effects of numbers of observations and predictors for various model types on the performance of forest inventory with airborne laser scanning

Author

Cosenza, Diogo N., Packalen, Petteri, Maltamo, Matti, Varvia, Petri, Raty, Janne, Soares, Paula, Tome, Margarida, Strunk, Jacob L., and Korhonen, Lauri

Subjects
Remote sensing -- Usage, Machine learning -- Usage, Forest management -- Technology application, Technology application, Earth sciences
Abstract

Semi- and nonparametric models are popular in the area-based approach (ABA) using airborne laser scanning. It is unclear, however, how many predictors and training plots are needed to provide accurate predictions without overfitting. This work aims to explore these limits for various approaches: ordinary least squares regression (OLS), generalized additive models (GAM), least absolute shrinkage and selection operator (LASSO), random forest (RF), support vector machine (SVM), and Gaussian process regression (GPR). We modeled timber volume ([m.sup.3]*[ha.sup.-1]) for four boreal sites using ABA with 2-39 predictors and 20-500 training plots. OLS, GAM, LASSO, and SVM overfitted as the number of predictors approached the number of training plots. They required [greater than or equal to] 15 plots per predictor to provide accurate predictions (RMSE [less than or equal to] 30%). GAM required [greater than or equal to] 250 plots regardless of the number of predictors. The number of predictors only mildly affected RF and GPR, but they required [greater than or equal to] 200 and [greater than or equal to] 250 training plots, respectively. RF did not overfit in any circumstances, whereas GPR overfit even with 500 training plots. Overall, using up to 39 predictors did not generally result in overfit, and for most model types, it resulted in better accuracy for sufficiently large datasets ([greater than or equal to] 250 plots). Key words: LiDAR, machine learning, remote sensing, area-based approach, sampling size. Les modeles semi et non parametriques sont populaires avec l'approche territoriale (AT) qui utilise le balayage laser aeroporte. On ignore cependant combien de predicteurs et de parcelles d'entrainement sont necessaires pour generer des predictions justes sans surapprentissage. Cet article vise a explorer ces limites pour differentes approches : regression des moindres carres ordinaire (MCO), modeles additifs generaux (MAG), operateur de selection et reduction par moindres valeurs absolues (LASSO); foret aleatoire (FA), machine a vecteurs de support (MVS), et regression de processus gaussien (RPG). Nous avonsmodelise le volume de bois ([m.sup.3]*[ha.sup.-1]) dans quatre stations boreales a l'aide de l'AT avec 2 a 39 predicteurs et 20 a 500 parcelles d'entrainement. Les approches MCO, MAG, LASSO et MVS surajustaient lorsque le nombre de predicteurs approchait le nombre de parcelles d'entrainement. Elles exigeaient au moins 15 parcelles par predicteur pour generer des predictions justes (EMQ [less than or equal to] 30%). L'approche MAG necessitait au moins 250 parcelles peu importe le nombre de predicteurs. Le nombre de predicteurs influencait seulement legerement les approches FA et RPG mais elles exigeaient respectivement au moins 200 et 250 parcelles. L'approche FA ne surajustait pas quelles que soient les circonstances, tandis que l'approche RPG surajustait meme avec 500 parcelles d'entrainement. Dans l'ensemble, l'utilisation de 39 predicteurs n'engendrait generalement pas de surapprentissage et, avec la plupart des types de modeles, cela se traduisait par une precision accrue pour des jeux de donnees suffisamment importants (au moins 250 parcelles). [Traduit par la Redaction] Mots-cles : LiDAR, apprentissage machine, teledetection, approche territoriale, taille de l'echantillon., 1. Introduction The development of prediction models is an essential step when performing forest assessments through airborne laser scanning (ALS). In the area-based approach (ABA) to forest inventory with ALS [...]

Published
2022
Full Text
View/download PDF

4. Gaussian process regression for forest attribute estimation from airborne laser scanning data

Author

Varvia, Petri, Lähivaara, Timo, Maltamo, Matti, Packalen, Petteri, and Seppänen, Aku

Subjects
Statistics - Applications
Abstract

While the analysis of airborne laser scanning (ALS) data often provides reliable estimates for certain forest stand attributes -- such as total volume or basal area -- there is still room for improvement, especially in estimating species-specific attributes. Moreover, while information on the estimate uncertainty would be useful in various economic and environmental analyses on forests, a computationally feasible framework for uncertainty quantifying in ALS is still missing. In this article, the species-specific stand attribute estimation and uncertainty quantification (UQ) is approached using Gaussian process regression (GPR), which is a nonlinear and nonparametric machine learning method. Multiple species-specific stand attributes are estimated simultaneously: tree height, stem diameter, stem number, basal area, and stem volume. The cross-validation results show that GPR yields on average an improvement of 4.6\% in estimate RMSE over a state-of-the-art k-nearest neighbors (kNN) implementation, negligible bias and well performing UQ (credible intervals), while being computationally fast. The performance advantage over kNN and the feasibility of credible intervals persists even when smaller training sets are used., Comment: Accepted for publication in IEEE Transactions on Geoscience and Remote Sensing

Published
2018
Full Text
View/download PDF

6. Three-Phase Hierarchical Model-Based and Hybrid Inference

Author

Saarela, Svetlana, primary, Varvia, Petri, additional, Korhonen, Lauri, additional, Yang, Zhiqiang, additional, Patterson, Paul L., additional, Gobakken, Terje, additional, Næsset, Erik, additional, Healey, Sean P., additional, and Ståhl, Göran, additional

Published
2023
Full Text
View/download PDF

14. Metsän heijastusmallin tilastollinen inversio

Author

Varvia, Petri, Sovelletun fysiikan laitos, Department of Applied Physics, Luonnontieteiden ja metsätieteiden tiedekunta, Sovelletun fysiikan laitos, Faculty of Science and Forestry, Department of Applied Physics, Luonnontieteiden ja metsätieteiden tiedekunta, and Faculty of Science and Forestry

Subjects
fysiikka, physics
Published
2013

15. Bayesian inversion of a forest reflectance model using Sentinel-2 and Landsat 8 satellite images

Author

Schraik, Daniel, Varvia, Petri, Korhonen, Lauri, Rautiainen, Miina, Department of Built Environment, Tampere University, University of Eastern Finland, Department of Electronics and Nanoengineering, Aalto-yliopisto, Aalto University, and Computing Sciences

Subjects
Clumping, Landsat 8, Spectral invariants, LIGHT-INTERCEPTION, LEAF-AREA INDEX, Bayesian inversion, 113 Computer and information sciences, PHOTON RECOLLISION PROBABILITY, LAI, Recollision probability, ANGLE DISTRIBUTION, Leaf area index, Forest reflectance, VEGETATION, Sentinel-2, CANOPY REFLECTANCE, PARAS
Abstract

openaire: EC/H2020/771049/EU//FREEDLES The inversion of reflectance models is a generalizable tool to obtain estimates on forest biophysical parameters, such as leaf area index, with theoretically little information need from a study area, instead relying on the knowledge about physical processes in the forest radiation regime. The use of prior information can greatly improve the reflectance model inversion, however, the literature does not yet provide much information on the selection of priors and their influence on the inversion results. In this study, we used a Bayesian approach to invert the PARAS forest reflectance model and retrieve leaf area index from Sentinel-2 MSI and Landsat 8 OLI multispectral satellite images. The PARAS model is based on the theory of spectral invariants, which describes the influence of wavelength-independent parameters on forest radiative transfer. The Bayesian inversion approach is highly flexible, provides uncertainty quantification, and enables the explicit incorporation of prior knowledge into the inversion process. We found that the choice of prior information is crucial in inverting a forest reflectance model to predict leaf area index. Regularizing and informative priors for leaf area index strongly improved the predictions, relative to an uninformative prior, in that they counteracted the saturation effect of the optical signal occuring at high values for leaf area index. The predictions of leaf area index were more accurate for Landsat 8 than for Sentinel-2, due to potential inconsistencies in the visible bands of Sentinel-2 in our data, and the higher spectral resolution. (C) 2019 The Authors. Published by Elsevier Ltd.

Full Text
View/download PDF

Catalog

Books, media, physical & digital resources
See catalog results