Search

To improve the design-based precision of gross increment estimates from forest inventories, we propose an assessment of an expanded set of auxiliary information grouped from five sources: (i) a vegetation height model, (ii) satellite imagery, (iii) spatial data, (iv) topography, and (v) variables identified by external forest monitoring and research networks. The former two are from optical remote sensing and the latter three are chosen on the basis of interpretable and proven connections with forest growth. We evaluate each source individually and collectively for the Swiss National Forest Inventory using two-phase estimation with the elastic net method. In terms of relative efficiency, all individual groups demonstrated improvement over one-phase estimation by 7% to 29% with the interpretable sources consistently outperforming those based on optical remote sensing. However, the interpretable sources do not provide significant additional gains when combined together, whereas the optical remote sensing consistently and substantially supports other sources of auxiliary data when combined, leading to a 50% to 71% improvement overall. Given the availability of data, such as that from international monitoring programs, expanding the set of auxiliaries to include both interpretable sources and optical remote sensing is a feasible and promising option for national forest inventories. [ABSTRACT FROM AUTHOR]

A simple design-based approach to estimating net change of a forest attribute such as timber volume is to observe the change directly on the plot level and then make that the response variable of interest using established estimation techniques such as multiphase regression estimators. This direct approach is only possible for inventories with permanent plots and is constrained to estimating net change over time periods matching the duration of the remeasurement cycle. Indirect estimation involves applying one of the aforementioned techniques to estimate the state at two time points and taking their difference. Indirect methods, although less common, are not necessarily constrained to permanent plots and can estimate net change over any desired time span for annual designs. This article compares design-based direct and indirect regression estimators under the Monte Carlo approach and illustrates their performances with data from the Swiss National Forest Inventory. The major finding is that direct estimation should be preferred whenever change is observable directly on the plot level but that multiphase indirect estimation can still improve precision when direct estimation is not possible such as for inventories employing only temporary plots. [ABSTRACT FROM AUTHOR]

This paper compares design-based properties of the classical two-phase regression estimator with several nonparametric kernel-based estimators of which k nearest neighbors ( kNN) is a special case. Metrics are based on the Euclidean distance applied to either a multidimensional space of explanatory variables or to a one-dimensional space of predictions obtained from a linear model. The main concepts of kernel-based regression estimators are reformulated in the design-based Monte Carlo approach to forest inventory. The results, based on a case study of a forest inventory in Switzerland and extensive simulations, suggest that the commonly used analytical external variance formula may systematically underestimate the true variance for a variety of kernel-based estimators including kNN but that it is still adequate for the classical regression estimator. Although using a bootstrap variance can help to correct this underestimation, it was also found that the bootstrap variance estimates could be unstable if the optimal bandwidth is recalculated in each bootstrap sample. These findings suggest that if the model captures the main features of the underlying process, then it is advisable to use the classical regression estimator, because it performs at least as well as the other techniques and is by far simpler to implement. [ABSTRACT FROM AUTHOR]

In 2009, the Swiss National Forest Inventory (NFI) turned from a periodic into an annual measurement design in which only one-ninth of the overall sample of permanent plots is measured every year. The reduction in sample size due to the implementation of the annual design results in an unacceptably large increase in variance when using the standard simple random sampling estimator. Thus, a flexible estimation procedure using two- and three-phase regression estimators is presented with a special focus on utilizing updating techniques to account for disturbances and growth and is applied to the second and third Swiss NFIs. The first phase consists of a dense sample of systematically distributed plots on a 500m x 500mgrid for which auxiliary variables are obtained through the interpretation of aerial photographs. The second phase is an eightfold looser subgrid with terrestrial plot data collected from the past inventory, and the third and final phase consists of the three most recent annual subgrids with the current state of the target variable (stem volume). The proposed three-phase estimators reduce the increase in variance from 294% to 145% compared with the estimator based on the full periodic sample while remaining unbiased. [ABSTRACT FROM AUTHOR]