scholarly journals Evaluating k-Nearest Neighbor (kNN) Imputation Models for Species-Level Aboveground Forest Biomass Mapping in Northeast China

2019 ◽  
Vol 11 (17) ◽  
pp. 2005 ◽  
Author(s):  
Yuanyuan Fu ◽  
Hong S. He ◽  
Todd J. Hawbaker ◽  
Paul D. Henne ◽  
Zhiliang Zhu ◽  
...  
Keyword(s):  
Forest Inventory ◽  
Environmental Variables ◽  
Northeast China ◽  
Nearest Neighbor ◽  
Forest Biomass ◽  
Remote Sensing Data ◽  
Species Level ◽  
Estimation Accuracy ◽  
Inventory Data ◽  
K Nearest Neighbor

Quantifying spatially explicit or pixel-level aboveground forest biomass (AFB) across large regions is critical for measuring forest carbon sequestration capacity, assessing forest carbon balance, and revealing changes in the structure and function of forest ecosystems. When AFB is measured at the species level using widely available remote sensing data, regional changes in forest composition can readily be monitored. In this study, wall-to-wall maps of species-level AFB were generated for forests in Northeast China by integrating forest inventory data with Moderate Resolution Imaging Spectroradiometer (MODIS) images and environmental variables through applying the optimal k-nearest neighbor (kNN) imputation model. By comparing the prediction accuracy of 630 kNN models, we found that the models with random forest (RF) as the distance metric showed the highest accuracy. Compared to the use of single-month MODIS data for September, there was no appreciable improvement for the estimation accuracy of species-level AFB by using multi-month MODIS data. When k > 7, the accuracy improvement of the RF-based kNN models using the single MODIS predictors for September was essentially negligible. Therefore, the kNN model using the RF distance metric, single-month (September) MODIS predictors and k = 7 was the optimal model to impute the species-level AFB for entire Northeast China. Our imputation results showed that average AFB of all species over Northeast China was 101.98 Mg/ha around 2000. Among 17 widespread species, larch was most dominant, with the largest AFB (20.88 Mg/ha), followed by white birch (13.84 Mg/ha). Amur corktree and willow had low AFB (0.91 and 0.96 Mg/ha, respectively). Environmental variables (e.g., climate and topography) had strong relationships with species-level AFB. By integrating forest inventory data and remote sensing data with complete spatial coverage using the optimal kNN model, we successfully mapped the AFB distribution of the 17 tree species over Northeast China. We also evaluated the accuracy of AFB at different spatial scales. The AFB estimation accuracy significantly improved from stand level up to the ecotype level, indicating that the AFB maps generated from this study are more suitable to apply to forest ecosystem models (e.g., LINKAGES) which require species-level attributes at the ecotype scale.

scholarly journals Landscape-scale parameterization of a tree-level forest growth model: a k-nearest neighbor imputation approach incorporating LiDAR data

2010 ◽  
Vol 40 (2) ◽  
pp. 184-199 ◽  
Author(s):  
Michael J. Falkowski ◽  
Andrew T. Hudak ◽  
Nicholas L. Crookston ◽  
Paul E. Gessler ◽  
Edward H. Uebler ◽  
...  
Keyword(s):  
Forest Inventory ◽  
Nearest Neighbor ◽  
Forest Growth ◽  
Tree Level ◽  
Lidar Data ◽  
Inventory Data ◽  
K Nearest Neighbor ◽  
Forest Inventory Data ◽  
Forest Growth Model ◽  
Nearest Neighbor Imputation

Sustainable forest management requires timely, detailed forest inventory data across large areas, which is difficult to obtain via traditional forest inventory techniques. This study evaluated k-nearest neighbor imputation models incorporating LiDAR data to predict tree-level inventory data (individual tree height, diameter at breast height, and species) across a 12 100 ha study area in northeastern Oregon, USA. The primary objective was to provide spatially explicit data to parameterize the Forest Vegetation Simulator, a tree-level forest growth model. The final imputation model utilized LiDAR-derived height measurements and topographic variables to spatially predict tree-level forest inventory data. When compared with an independent data set, the accuracy of forest inventory metrics was high; the root mean square difference of imputed basal area and stem volume estimates were 5 m2·ha–1 and 16 m3·ha–1, respectively. However, the error of imputed forest inventory metrics incorporating small trees (e.g., quadratic mean diameter, tree density) was considerably higher. Forest Vegetation Simulator growth projections based upon imputed forest inventory data follow trends similar to growth projections based upon independent inventory data. This study represents a significant improvement in our capabilities to predict detailed, tree-level forest inventory data across large areas, which could ultimately lead to more informed forest management practices and policies.

K Nearest Neighbor Method for Forest Inventory Using Remote Sensing Data

2007 ◽  
Vol 44 (2) ◽  
pp. 149-165 ◽  
Author(s):  
Qingmin Meng ◽  
Chris J. Cieszewski ◽  
Marguerite Madden ◽  
Bruce E. Borders
Keyword(s):  
Remote Sensing ◽  
Forest Inventory ◽  
Nearest Neighbor ◽  
Remote Sensing Data ◽  
K Nearest Neighbor ◽  
Sensing Data

scholarly journals Mapping aboveground biomass by integrating geospatial and forest inventory data through a k-nearest neighbor strategy in North Central Mexico

2013 ◽  
Vol 6 (1) ◽  
pp. 80-96 ◽  
Author(s):  
Carlos A. Aguirre-Salado ◽  
Eduardo J. Treviño-Garza ◽  
Oscar A. Aguirre-Calderón ◽  
Javier Jiménez-Pérez ◽  
Marco A. González-Tagle ◽  
...  
Keyword(s):  
Aboveground Biomass ◽  
Forest Inventory ◽  
Nearest Neighbor ◽  
Central Mexico ◽  
Inventory Data ◽  
K Nearest Neighbor ◽  
North Central ◽  
Forest Inventory Data

Forest Biomass Estimation Based on Forest Inventory Data in Middle and Last Scales

2011 ◽  
Vol 183-185 ◽  
pp. 220-224
Author(s):  
Ming Ze Li ◽  
Wen Yi Fan ◽  
Ying Yu
Keyword(s):  
Tree Species ◽  
Forest Inventory ◽  
Forest Biomass ◽  
Biomass Estimation ◽  
Total Biomass ◽  
Inventory Data ◽  
Biomass Equation ◽  
Tree Species Composition ◽  
Forest Inventory Data ◽  
Ecosystem Carbon

The forest biomass (which is referred to the arbor aboveground biomass in this research) is one of the most primary factors to determine the forest ecosystem carbon storages. There are many kinds of estimating methods adapted to various scales. It is a suitable method to estimate forest biomass of the farm or the forestry bureau in middle and last scales. First each subcompartment forest biomass should be estimated, and then the farm or the forestry bureau forest biomass was estimated. In this research, based on maoershan farm region, first the single tree biomass equation of main tree species was established or collected. The biomass of each specie was calculated according to the materials of tally, such as height, diameter and so on in the forest inventory data. Secondly, each specie’s biomass and total biomass in subcompartment were calculated according to the tree species composition in forest management investigation data. Thus the forest biomass spatial distribution was obtained by taking subcompartment as a unit. And last the forest total biomass was estimated.

Estimation of the North–South Transect of Eastern China forest biomass using remote sensing and forest inventory data

2013 ◽  
Vol 34 (15) ◽  
pp. 5598-5610 ◽  
Author(s):  
Yanhua Gao ◽  
Xinxin Liu ◽  
Chengcheng Min ◽  
Honglin He ◽  
Guirui Yu ◽  
...  
Keyword(s):  
Remote Sensing ◽  
Forest Inventory ◽  
Eastern China ◽  
Forest Biomass ◽  
Inventory Data ◽  
Forest Inventory Data ◽  
The North

scholarly journals Flood Detection and Susceptibility Mapping Using Sentinel-1 Remote Sensing Data and a Machine Learning Approach: Hybrid Intelligence of Bagging Ensemble Based on K-Nearest Neighbor Classifier

2020 ◽  
Vol 12 (2) ◽  
pp. 266 ◽  
Author(s):  
Himan Shahabi ◽  
Ataollah Shirzadi ◽  
Kayvan Ghaderi ◽  
Ebrahim Omidvar ◽  
Nadhir Al-Ansari ◽  
...  
Keyword(s):  
Nearest Neighbor ◽  
Statistical Error ◽  
Remote Sensing Data ◽  
Susceptibility Mapping ◽  
Training Dataset ◽  
Mapping Technique ◽  
K Nearest Neighbor ◽  
Northern Iran ◽  
Ensemble Models ◽  
Flood Prone Areas

Mapping flood-prone areas is a key activity in flood disaster management. In this paper, we propose a new flood susceptibility mapping technique. We employ new ensemble models based on bagging as a meta-classifier and K-Nearest Neighbor (KNN) coarse, cosine, cubic, and weighted base classifiers to spatially forecast flooding in the Haraz watershed in northern Iran. We identified flood-prone areas using data from Sentinel-1 sensor. We then selected 10 conditioning factors to spatially predict floods and assess their predictive power using the Relief Attribute Evaluation (RFAE) method. Model validation was performed using two statistical error indices and the area under the curve (AUC). Our results show that the Bagging–Cubic–KNN ensemble model outperformed other ensemble models. It decreased the overfitting and variance problems in the training dataset and enhanced the prediction accuracy of the Cubic–KNN model (AUC=0.660). We therefore recommend that the Bagging–Cubic–KNN model be more widely applied for the sustainable management of flood-prone areas.

Economic losses in carbon forestry due to errors in inventory data

2020 ◽  
Author(s):  
Roope Ruotsalainen ◽  
Timo Pukkala ◽  
Annika Kangas ◽  
Mari Myllymäki ◽  
Petteri Packalen
Keyword(s):  
Forest Inventory ◽  
Laser Scanning ◽  
Net Present Value ◽  
Nearest Neighbor ◽  
Carbon Price ◽  
Wood Products ◽  
Total Carbon ◽  
Economic Losses ◽  
Inventory Data ◽  
Forest Inventory Data

Forestry can help to mitigate climate change by storing carbon in trees, forest soils and wood products. Forest owners can be subsidized if forestry removes carbon from the atmosphere and taxed if forestry produces emissions. Errors in forest inventory data can lead to losses in net present value (NPV) if management prescriptions are selected based on erroneous data but not on correct data. This study assesses the effect of inventory errors on economic losses in forest management when the objective is to maximize the total NPV of timber production and carbon payments. Errors similar as in airborne laser scanning based forest inventory were simulated in stand attributes with a vine copula approach and nearest neighbor method. Carbon payments were based on the total carbon balance of forestry (incl. trees, soil and wood-based products) and calculations were carried out for 30 years using carbon prices of € 0, 50, 75, 100, 125 and 150 t-1. The results revealed that increasing the carbon price and decreasing the level of errors led to decreased losses in NPV. The inclusion of carbon payments for the maximization of the NPV decreased the effect of errors on the losses, which suggests that the value of collecting more accurate forest inventory data may decrease when the carbon price increases.

Sign in / Sign up

Export Citation Format

Share Document