Evaluating Site-Specific and Generic Spatial Models of Aboveground Forest Biomass Based on Landsat Time-Series and LiDAR Strip Samples in the Eastern USA

Abstract

Large-area assessment of aboveground tree biomass (AGB) to inform regional or national forest monitoring programs can be efficiently carried out by combining remotely sensed data and field sample measurements through a generic statistical model, in contrast to site-specific models. We integrated forest inventory plot data with spatial predictors from Landsat time-series imagery and LiDAR strip samples at four sites across the eastern USA-Minnesota (MN), Maine (ME), Pennsylvania-New Jersey (PANJ) and South Carolina (SC)-in statistical modeling frameworks to analyze the performance of generic (all sites combined) versus site-specific models. The major objective was to evaluate the prediction accuracy of generic and site-specific models when applied to particular sites. Pixel-level polynomial model fitting was applied to the time-series of near-anniversary date Landsat variables to obtain projected metrics in the target year 2014 for which LiDAR strip samples were available. Two forms of models based on ordinary least-squares multiple linear regressions (MLR) and the random forest (RF) machine learning approach were developed for each site and for the pooled (i.e., generic) reference data frame. The models were evaluated using national forest inventory (NFI) data for the USA. We observed stronger fit statistics with the MLR than with RF for both the site-specific and the generic models. The proportions of variances explained (adjusted R 2 ) with the site-specific models were 0.86, 0.78, 0.82 and 0.92 for ME, MN, PANJ and SC, respectively while the generic model had adjusted R 2 = 0.85. A test of statistical equivalence of observed and predicted AGB for the NFI locations did not reveal equivalence with any of the models, possibly due to the different resolutions of the observed and predicted data. In contrast, predictions by the generic and site-specific models were equivalent. We conclude that a generic model provides accuracies comparable to the site-specific models for large-area AGB assessment across our study sites in the eastern USA.

Using matrix models to estimate aboveground forest biomass dynamics in the eastern USA through various combinations of LiDAR, Landsat, and forest inventory data

“…Among the selected variables from both LiDAR and Landsat data, LiDAR metrics were much more influential to stand basal area and associated AGB than Landsat variables. This finding is consistent with previous studies (Deo et al 2017a(Deo et al , 2017b. Nevertheless, we found that Landsat variables alone can also be used to replace stand basal area to predict AGB in the study region although with higher RMSE (9.34-12 Mg ha −1 ).…”

“…The average width of the strips was about 1.2 km. The LiDAR datasets were summarized to obtain the grid metrics for canopy cover and vertical strata density as defined in Deo et al (2017aDeo et al ( , 2017b. In addition, the Landsat time-series data were acquired by Landsat-5 Thematic Mapper (TM) and Landsat-8 Operational Land Imager (OLI) sensors at the peak of the growing seasons (July-September) from 1984-2015.…”

“…Heath et al 2011, Woodall et al 2015a. Recent work has highlighted that the accuracy and precision of AGB dynamics over a range of spatial scales could be further improved by incorporating vegetation indices as predictors derived from remotely sensed data that are temporally consistent and spatially continuous (Deo et al 2017a). Providing spatially continuous and temporally consistent estimates of AGB dynamics in forests at large-scales is crucial for evaluating existing land use policies and land management practices intended to help mitigate GHG emissions (Woodall et al 2015b, Deo et al 2017b.…”

Automatic Detection of Single Trees in Airborne Laser Scanning Data through Gradient Orientation Clustering

“…Airborne laser scanning (ALS) is a widely used LiDAR technology in forestry, which has a unique advantage in 3D terrain and forest height detection. It can quickly and accurately obtain a digital surface model (DSM) and forest height information for forested land [4,5]. According to the statistics, single-tree detection algorithms that use active RS data, such as ALS, have grown rapidly in recent years, accounting for 80% of the total increase in RS area, at the same time, those algorithms using passive data or a fusion of passive and active data accounts for 8% and 12% respectively [6].…”

Estimation of Forest Biomass Patterns across Northeast China Based on Allometric Scale Relationship

“…As the principal part of terrestrial ecosystems, forest ecosystems hold approximately 80% of the terrestrial aboveground and below-ground biomass, and play very important roles in the global carbon cycle and climate change [1][2][3]. Since the temperate and boreal forests play a crucial role as atmospheric CO2 sinks, more attention has been paid to climatic warming in mid-and high-latitudes…”