Search

Fiscal uncertainties can sometimes affect national continuous forest monitoring efforts. One solution of interest is to lengthen the time it takes to collect a "full set" of plot data from five to 10 years in order to reduce costs. Here, we investigate using ancillary information to partially offset this proposed solution's negative effects. We focus our discussion on the corresponding number of years between measurements of each plot while we investigate how thoroughly the detrimental effects of the reduced sampling effort can be ameliorated with change estimates obtained from temporally-dense remotely-sensed images. We simulate measured plot data under four sampling error structures, and we simulate remotely-sensed change estimates under three reliability assumptions, integrated with assumptions about the additional unobserved growth resulting from the lengthened observation window. We investigate a number of estimation systems with respect to their ability to provide compatible annual estimates of the components of change during years spanned by at least half of the full set of plot observations. We show that auxiliary data with shorter observation intervals can contribute to a significant improvement in estimation. [ABSTRACT FROM AUTHOR]

Continuous forest inventory planners can allocate the budget to more plots per acre or a shorter remeasurement cycle. A higher plot intensity benefits small area estimation and allows for more precision in current status estimates. Shorter cycles may provide better estimates of growth, removals and mortality. On a fixed budget, the planner can't have both greater plot intensity and shorter cycles. Therefore, it is important to understand the trade-offs involved. Growth over removals ratios are important indicators of sustainability, and can be adversely affected by changes in cycle length. However, it might be possible to ameliorate negative impacts of longer cycle lengths with judicious use of aerial imagery. Increasing the cycle length reduces the value of an inventory for monitoring, but reducing the number of plots increases the variance of both current status estimates and trend estimates. There may be no optimal statistical solution to this quandary, but the best solution will depend on policy and management objectives. Continuous forest inventories use permanent plots that are remeasured to provide information on growth, removals and mortality. Typically, all plots are remeasured within a narrow time span, but the USDA Forest Service has popularized a variant referred to as an annual forest inventory where a percentage of the permanent plots are remeasured every year. We discuss trade-offs between number of field plots and cycle length and provide some insight with example applications showing how these decisions impact growth and removals estimates. We also discuss a variant of the traditional growth over removals ratio estimator that limits degradation in estimate quality as cycle lengths increase. [ABSTRACT FROM AUTHOR]

Historically, the goal of forest inventories has been to determine the extent of the timber resource. Predictions of how the resource was changing were made by comparing differences between successive inventories. The general view of the associated sample design was with selection probabilities based on land area observed at a discrete point in time. Time was not considered to be part of the sample design because it was not considered to be an element of the sampled population. Over the last few decades, the general goal of many national-scale forest inventories has been changing to monitoring the dynamic forest ecosystem. Here, we explore the inferential advantages of replacing the two-dimensional areal probability paradigm with a three-dimensional spatiotemporal probability paradigm. Our general discussion is augmented with a simulated example for estimating annual growth by diameter classes. Two assumptions of temporal indifference and remeasurement interval length indifference, which arose because of the two-dimensional view, are investigated through a simulation. The simulation compares and contrasts five estimators that differ in their reliance on those assumptions to make annual estimates. The results of the simulations often show those assumptions to be bias inducing. [ABSTRACT FROM AUTHOR]

The coarsened exact matching (CEM) method is used to match annual forest inventory plots awaiting remeasurement with plots that have already been remeasured. This results in a model-free approach for short term inventory projections. CEM has many desirable properties relative to other matching methods and is easy to apply within a SQL database. The combination of short term projections with a 3 or 5 year moving window is suggested for providing trend estimates that include the current year and a few years into the future. The default projection represents business as usual. A method to bias the plot matching to generate desired scenarios is also developed. These ideas and methods are demonstrated with several applications to forest inventory data. Scenarios are generated where increasing future harvest levels are stochastically controlled to demonstrate this capability with operational data. [ABSTRACT FROM AUTHOR]

Simple methods for estimating the proportion of land changing from forest to nonforest are developed. Variance estimators are derived to facilitate significance tests. A power analysis indicates that 400 inventory plots are required to reliably detect small changes in net or gross forest loss. This is an important result because forest certification programs may require additional precautions when wood from areas where forestland area loss is occurring is harvested or purchased. Net and gross forest area loss must be clearly differentiated to avoid confusion. Estimates of gross forest cover loss from satellite data should not be equated with net forest area loss, which can be better determined from remeasured forest inventory plots. Simultaneous tests of net and gross forest area loss should use multiple comparison procedures to ensure that overall error rates are correct. Examples of applications demonstrate how to properly perform these tests. A simulated example is used to verify that the variance estimators are reliable. An application to USDA Forest Service inventory data indicates that neither net nor gross forest loss at the state level was statistically significant for states that had sufficien remeasured plot data publicly available when this analysis was done. [ABSTRACT FROM AUTHOR]

Wildlife studies to determine preferred habitat often involve sampling used sites and making comparisons with unused background sites. We apply these data to a somewhat different purpose, which is to assess the spatial distribution of important habitat variables relative to spotted owl nest sites. It is necessary to compare habitat characteristics of used and unused sites to understand habitat preferences. However, ecological literature provides little guidance for selection of appropriate background sites necessary for such an analysis. We apply propensity score matching to telemetry gathered samples of spotted owl habitat for the purpose of weighting the selected background sites. This helps to ensure that important covariates have similar distributions in the used and unused habitat samples. We propose a two-stage process where propensity score matching of used and unused sites results in weights that are applied in the second stage modelling process. The objective is to quantify how wildlife responses to important variables change with distance from a nest, den site or other central location. The expectation is that this will strengthen guidance to managers who wish to protect or to improve wildlife habitat on their forest land. Example applications using spotted owl telemetry data demonstrate how to apply the methods and show the potential to elucidate habitat characteristics. [ABSTRACT FROM AUTHOR]

The transitioning of land from forest to other uses is of increasing interest as urban areas expand and the world's population continues to grow. Also of interest, but less recognized, is the transitioning of land from other uses into forest. In this paper, we show how rates of conversion from forest to non-forest and non-forest to forest can be estimated in the US from a continuously improving publicly available annual forest inventory database, under varying definitions of conversion. Two estimation approaches are considered and contrasted. The approaches are a simple ratio estimator and the weighted maximum likelihood estimator. The latter involves a statistical procedure that incorporates the binomial nature of the indicator variables, the transition of mapped plot conditions and an intuitively appealing way to combine data from varying remeasurement periods for a temporally dependent variable. [ABSTRACT FROM PUBLISHER]

Line transects are commonly used for sampling coarse woody debris (CWD). The USDA Forest Service Forest Inventory and Analysis programme uses a variant of this method that involves sampling for CWD along transects that radiate from the centre of a circular plot-like spokes on a wheel. A new approach for analysis of data collected with spoked transects is developed. This approach does not attempt to mimic traditional analysis methods for line transects, which do not readily apply to spoked transects. The method developed here generalizes to any number of spokes. We also demonstrate that there is a close relationship between spoked transects and a particular fixed plot design for sampling CWD. [ABSTRACT FROM AUTHOR]

The use of fixed-area plots for sampling down coarse woody debris is reviewed. A set of clearly defined protocols for two previously described methods is established and a new method, which we call the ‘sausage’ method, is developed. All methods (protocols) are shown to be unbiased for volume estimation, but not necessarily for estimation of population density in terms of number of down pieces. The new sausage method performs best overall in terms of minimum variance criteria. This review should also help establish a description of the major fixed-area plot sampling methods that will allow unambiguous application and description in future studies. [ABSTRACT FROM AUTHOR]

Managing the forest to store carbon is a relatively new concept. Various regional greenhouse gas initiatives and new Federal legislation are providing financial incentives for forest owners to manage for carbon in addition to other forest products. These incentives are intended for landowners who engage in activities that go beyond business as usual practices. Managing for carbon will likely involve foregoing other investment alternatives and increasing rotation lengths. The analysis approach demonstrated here provides a relatively simple method for an owner to compare traditional forest management and regular harvests with letting the trees grow to accumulate more carbon in the forest. Several financial decision statistics are considered and demonstrated with examples. A derivative of land expectation value, called rotation equivalent value, is shown to be a useful decision tool for comparing carbon storage with other management options having different rotation lengths. L'aménagement de la forêt pour emmagasiner du carbone est un concept relativement nouveau. Diverses initiatives régionales portant sur les gaz à effet de serre et une nouvelle législation fédérale incitent les propriétaires de forêts à aménager en fonction du carbone en plus des autres produits forestiers. Ces mesures incitatives s'adressent aux propriétaires terriens qui s'engagent dans des activités qui vont au-delà des pratiques traditionnelles. Aménager pour le carbone impliquera probablement de laisser de côté d'autres choix d'investissement et d'allonger la durée des rotations. L'approche d'analyse présentée ici fournit une méthode relativement simple pour qu'un propriétaire puisse comparer l'aménagement forestier traditionnel et les coupes régulières avec le fait de laisser croître les arbres pour accumuler plus de carbone dans la forêt. Plusieurs statistiques associées à la prise de décisions financières sont prises en compte et illustrées par des exemples. Une dérivée de la valeur d'attente du fonds, appelée valeur équivalente de rotation, s'avère un outil utile pour comparer l'emmagasinage du carbone avec d'autres options d'aménagement qui ont différentes durées de rotation. [ABSTRACT FROM AUTHOR]

Forest certification is designed to assure consumers that the wood products they purchase come from sustainably managed forests. Certified forests must meet certain standards that vary with the certification scheme. However, certification generally promotes environmental objectives by restricting forest management options in ways that may reduce wood and fibre production and increase the cost of delivering wood to market. For example, some aspects of certification programmes can discourage creation of new forest plantations and application of herbicide and fertilizer. These restrictions have costs which often are difficult to quantify. Bounds on the opportunity costs of restricting silvicultural options are explored by comparing a natural pine regime to plantation management regimes with and without herbicide using some basic economic criteria. In this analysis, a plantation regime with herbicide had nearly four times the land expectation value (LEV) of a natural pine regime and almost double the LEV of a plantation without herbicide treatment. [ABSTRACT FROM PUBLISHER]

Annual forest inventories present special challenges and opportunities for those analyzing the data arising from them. Here, we address one question currently being asked by analysts of the US Forest Service's Forest Inventory and Analysis Program's quickly accumulating annual inventory data. The question is simple but profound: When combining the next year's data for a particular variable with data from previous years, how does one know whether the same model as used in the past for this purpose continues to be applicable? Of the myriad approaches that have been developed for changepoint detection and anomaly detection, this report focuses on a simple quality-control approach known as a control chart that will allow analysts of annual forest inventory data to determine when a departure from a past trend is likely to have occurred. [ABSTRACT FROM AUTHOR]

In the past decade, the United States Department of Agriculture Forest Service's Forest Inventory and Analysis Program (FIA) has replaced regionally autonomous, periodic, state-wide forest inventories using various probability proportional to tree size sampling designs with a nationally consistent annual forest inventory design utilizing systematically spaced clusters of fixed-area plots. This has resulted in significant changes in the spatial-temporal distribution of observations on the nation's forest resource. This paper discusses the resulting changes in the observation of the distribution of tree diameters measured at 1.37 m above ground level. We show that in three of the four FIA regions, a significant portion of the upper end of the diameter distribution has thus far gone unobserved by the new inventory design. We conclude that the reason is not that the larger diameter trees have become less common but rather that there is a low enough probability of observing these trees that they are not being observed. This postulate is supported by the observation that large-diameter trees are not missing from the data acquired by the fourth FIA region (the Pacific Northwest), which uses an additional sampling mechanism for large-diameter trees. We explore the implications of this effect in addition to potential solutions. [ABSTRACT FROM PUBLISHER]

Mapped plots are used for the national forest inventory conducted by the US Forest Service. Mapped plots are also useful for single ownership inventories. Mapped plots can handle boundary overlap and can provide less variable estimates for specified forest conditions. Mapping is a good fit for fixed plot inventories where the fixed area plot is used for both mapping and tree sampling. However, mapping can be done with a colocated fixed area plot even though the tree inventory uses variable plots. In fact, the tree measurement plots can be completely decoupled from the plots used for measuring condition proportions, although this is not recommended. [ABSTRACT FROM AUTHOR]

A volume change index is suggested that is derived from growth components that can be estimated from remeasured plots. The new index incorporates more information than the traditional growth over removals ratio. The new index directly indicates whether the standing volume will be increasing or decreasing if current conditions persist, whereas the ratio of growth over removals obscures the impact of mortality. The details of estimating the components of growth are discussed with regard to interpreting the new index. The effect of spatial scale on the index is explored and a variance estimator is suggested. The new index is estimated from US Department of Agriculture Forest Service annual inventory data for two example applications. The first application looks at spatial trends in the index. The second application applies the index to a mill working circle. [ABSTRACT FROM AUTHOR]

Alternative definitions for growth and removals from a forest being monitored over time are discussed. It is shown that the definitions that are used in practice may not be what one would expect and estimates can vary substantially under alternative definitions. This can result in conclusions about forest sustainability that may be misleading. Alternative definitions are applied to selected states using US Department of Agriculture Forest Service inventory data. Standard errors of growth over removals ratios are used to indicate potential sustainability problems for selected forest types. [ABSTRACT FROM AUTHOR]

A region can conceptually be tessellated into polygons at different scales or resolutions. Likewise, samples can be taken from the region to determine the value of a polygon variable for each scale. Sampled polygons can be used to estimate values for other polygons at the same scale. However, estimates should be compatible across the different scales. Estimates are often required for zones within a region, where a region might be a state and counties could be the zones. A method is developed for estimating high-resolution (pixel) values that are constrained to be compatible with results for lower resolution values. The high-resolution values are constrained to sum to totals for zones within a region, where the totals are being simultaneously estimated from measurements taken at a different scale. If the zone estimates are unbiased, then the pixel-based estimates for the zone will be less biased. Sums of pixels in arbitrary polygons are thereby constrained to approach unbiased estimates. Approximate variance estimators are developed for the summed pixel estimates. Two example applications are provided. The first example is based on simulated data and verifies that the proposed variance estimators give reasonable results. The second example estimates the volume in a circle around a possible mill site in North Carolina. This example uses publicly available US Forest Service inventory data and simulated inventory data that the mill would provide. [ABSTRACT FROM AUTHOR]

Copyright of Canadian Journal of Forest Research is the property of Canadian Science Publishing and its content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written permission. However, users may print, download, or email articles for individual use. This abstract may be abridged. No warranty is given about the accuracy of the copy. Users should refer to the original published version of the material for the full abstract. (Copyright applies to all Abstracts.)

Copyright of Canadian Journal of Forest Research is the property of Canadian Science Publishing and its content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written permission. However, users may print, download, or email articles for individual use. This abstract may be abridged. No warranty is given about the accuracy of the copy. Users should refer to the original published version of the material for the full abstract. (Copyright applies to all Abstracts.)

Analysis of capture—recapture data often involves maximizing a complex likelihood function with many unknown parameters. Statistical inference based on selection of a proper model depends on successful attainment of this maximum. An EM algorithm is developed for obtaining maximum likelihood estimates of capture and survival probabilities conditional on first capture from standard capture—recapture data. The algorithm does not require the use of numerical derivatives which may improve precision and stability relative to other estimation schemes. The asymptotic covariance matrix of the estimated parameters can be obtained using the supplemented EM algorithm. The EM algorithm is compared to a more traditional Newton-Raphson algorithm with both a simulated and a real dataset. The two algorithms result in the same parameter estimates, but Newton-Raphson variance estimates depend on a numerically estimated Hessian matrix that is sensitive to step size choice. [ABSTRACT FROM AUTHOR]

Copyright of Canadian Journal of Forest Research is the property of Canadian Science Publishing and its content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written permission. However, users may print, download, or email articles for individual use. This abstract may be abridged. No warranty is given about the accuracy of the copy. Users should refer to the original published version of the material for the full abstract. (Copyright applies to all Abstracts.)

Copyright of Canadian Journal of Forest Research is the property of Canadian Science Publishing and its content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written permission. However, users may print, download, or email articles for individual use. This abstract may be abridged. No warranty is given about the accuracy of the copy. Users should refer to the original published version of the material for the full abstract. (Copyright applies to all Abstracts.)

Presents the comparison of three estimators of current status and trend for an annual interpenetrating panel design of forests. How the moving average tends to lag evolving trends; Findings of a large bias in quadratic growth trends.

Flexible models are proposed for assessing trend in annual forest survey data with increasing levels of generality. The models can be fitto the data using a mixed estimator and compared using the Akaike orSchwarz information criterion. The models also provide smoothed estimates of annual means. [ABSTRACT FROM AUTHOR]

Abstract unavailable [ABSTRACT FROM AUTHOR]

Annual measurement of a subset of permanent plots provides timely information on the forest resource. A variety of methods for analyzing the data from an annual survey are discussed, but coverage of all possible methods is not attempted. The methods considered range from application of standard systematic sampling procedures to more elaboratetime series methods. The bulk of the discussion concerns analysis procedures for singly and multiply imputed data sets, which contain imputed values for plots not measured during the current year. This is accompanied by a limited discussion of imputing methods. [ABSTRACT FROM AUTHOR]