Multi-source forest inventory data for forest production and utilization analyses at different levels

Haakana, Helena

doi:10.14214/df.243

Cited by 4 publications

(3 citation statements)

References 131 publications

(225 reference statements)

Supporting

Mentioning

Contrasting

Order By: Relevance

“…In situ measurements provide transparent, comparable and consistent information on the availability of forest resources for the development of forest-associated policies and to support decision making [5]. Combined with remote sensing instruments, the sparse network of field plots can provide analyses for larger areas [6]. Airborne LiDAR measurements and aerial imagery enable accurate forest inventory and monitoring, but airborne acquisitions are expensive, limited in spatial and temporal coverage, and can contain data gaps, thus restricting their utility for wall-to-wall mapping of large areas [7].…”

Section: Introductionmentioning

confidence: 99%

Deep Neural Networks with Transfer Learning for Forest Variable Estimation Using Sentinel-2 Imagery in Boreal Forest

et al. 2021

View full text Add to dashboard Cite

Estimation of forest structural variables is essential to provide relevant insights for public and private stakeholders in forestry and environmental sectors. Airborne light detection and ranging (LiDAR) enables accurate forest inventory, but it is expensive for large area analyses. Continuously increasing volume of open Earth Observation (EO) imagery from high-resolution (<30 m) satellites together with modern machine learning algorithms provide new prospects for spaceborne large area forest inventory. In this study, we investigated the capability of Sentinel-2 (S2) image and metadata, topography data, and canopy height model (CHM), as well as their combinations, to predict growing stock volume with deep neural networks (DNN) in four forestry districts in Central Finland. We focused on investigating the relevance of different input features, the effect of DNN depth, the amount of training data, and the size of image data sampling window to model prediction performance. We also studied model transfer between different silvicultural districts in Finland, with the objective to minimize the amount of new field data needed. We used forest inventory data provided by the Finnish Forest Centre for model training and performance evaluation. Leaving out CHM features, the model using RGB and NIR bands, the imaging and sun angles, and topography features as additional predictive variables obtained the best plot level accuracy (RMSE% = 42.6%, |BIAS%| = 0.8%). We found 3×3 pixels to be the optimal size for the sampling window, and two to three hidden layer DNNs to produce the best results with relatively small improvement to single hidden layer networks. Including CHM features with S2 data and additional features led to reduced relative RMSE (RMSE% = 28.6–30.7%) but increased the absolute value of relative bias (|BIAS%| = 0.9–4.0%). Transfer learning was found to be beneficial mainly with training data sets containing less than 250 field plots. The performance differences of DNN and random forest models were marginal. Our results contribute to improved structural variable estimation performance in boreal forests with the proposed image sampling and input feature concept.

show abstract

Section: Introductionmentioning

confidence: 99%

Deep Neural Networks with Transfer Learning for Forest Variable Estimation Using Sentinel-2 Imagery in Boreal Forest

et al. 2021

View full text Add to dashboard Cite

show abstract

“…They are available for some regions and useful in addressing forest owners', governmental, and independent organizations' needs [14]. However, these data do not cover all regions of practical interest [15]. Furthermore, such data actualization is time-consuming and cost intensive in difficult-to-access areas.…”

Section: Introductionmentioning

confidence: 99%

Estimation of the Canopy Height Model From Multispectral Satellite Imagery With Convolutional Neural Networks

et al. 2022

View full text Add to dashboard Cite

The canopy height model (CHM) is a representation of the height of the top of vegetation from the surrounding ground level. It is crucial for the extraction of various forest characteristics, for instance, timber stock estimations and forest growth measurements. There are different ways of obtaining the vegetation height, such as through ground-based observations or the interpretation of remote sensing images. The severe downside of field measurement is its cost and acquisition difficulty. Therefore, utilizing remote sensing data is, in many cases, preferable. The enormous advances in computer vision during the previous decades have provided various methods of satellite imagery analysis. In this work, we developed the canopy height evaluation workflow using only RGB and NIR (near-infrared) bands of a very high spatial resolution (investigated on WorldView-2 satellite bands). Leveraging typical data from airplane-based LiDAR (Light Detection and Ranging), we trained a deep neural network to predict the vegetation height. The provided approach is less expensive than the commonly used drone measurements, and the predictions have a higher spatial resolution (less than 5 m) than the vast majority of studies using satellite data (usually more than 30 m). The experiments, which were conducted in Russian boreal forests, demonstrated a strong correlation between the prediction and LiDAR-derived measurements. Moreover, we tested the generated CHM as a supplementary feature in the species classification task. Among different input data combinations and training approaches, we achieved the mean absolute error equal to 2.4 m using U-Net with Inception-ResNet-v2 encoder, high-resolution RGB image, near-infrared band, and ArcticDEM. The obtained results show promising opportunities for advanced forestry analysis and management. We also developed the easyto-use open-access solution for solving these tasks based on the approaches discussed in the study cloud-free composite orthophotomap provided by Mapbox via tile-based map service.

show abstract

“…In smaller scales, remote sensing is the only available data source supporting change monitoring. In addition, remote sensing forms an important resource in sustainable forest management, because it provides spatially explicit information on various forest attributes used as ecosystem service indicators (Haakana 2017;Vauhkonen and Ruotsalainen 2017). Czúcz and Condé (2018) list, among others, forest tree species richness, deadwood, forest age and growing stock as the key ecosystem condition indicators for forests; most of them can be assessed employing remote sensing.…”

Section: Introductionmentioning

confidence: 99%

Assessing local trends in indicators of ecosystem services with a time series of forest resource maps

Katila¹,

Rajala²,

Kangas³

2020

Silva Fenn.

View full text Add to dashboard Cite

Since the 1990âs, forest resource maps and small area estimates have been produced by combining national forest inventory (NFI) field plot data, optical satellite images and numerical map data using a non-parametric -nearest neighbour method. In Finland, thematic maps of forest variables have been produced by the means of multi-source NFI (MS-NFI) for eight to ten times depending on the geographical area, but the resulting time series have not been systematically utilized. The objective of this study was to explore the possibilities of the time series for monitoring the key ecosystem condition indicators for forests. To this end, a contextual Mann-Kendall (CMK) test was applied to detect trends in time-series of two decades of thematic maps. The usefulness of the observed trends may depend both on the scale of the phenomena themselves and the uncertainties involved in the maps. Thus, several spatial scales were tested: the MS-NFI maps at 16 Ã 16 m pixel size and units of 240 Ã 240 m, 1200 Ã 1200 m and 12â000 Ã 12â000 m aggregated from the MS-NFI map data. The CMK test detected areas of significant increasing trends of mean volume on both study sites and at various unit sizes except for the original thematic map pixel size. For other variables such as the mean volume of tree species groups, the proportion of broadleaved tree species and the stand age, significant trends were mostly found only for the largest unit size, 12â000 Ã 12â000 m. The multiple testing corrections decreased the amount of significant -values from the CMK test strongly. The study showed that significant trends can be detected enabling indicators of ecosystem services to be monitored from a time-series of satellite image-based thematic forest maps.k22222p

show abstract

Multi-source forest inventory data for forest production and utilization analyses at different levels

Cited by 4 publications

References 131 publications

Deep Neural Networks with Transfer Learning for Forest Variable Estimation Using Sentinel-2 Imagery in Boreal Forest

Deep Neural Networks with Transfer Learning for Forest Variable Estimation Using Sentinel-2 Imagery in Boreal Forest

Estimation of the Canopy Height Model From Multispectral Satellite Imagery With Convolutional Neural Networks

Assessing local trends in indicators of ecosystem services with a time series of forest resource maps

Contact Info

Product

Resources

About