A comparison of small-area estimation techniques to estimate selected stand attributes using LiDAR-derived auxiliary variables
نویسندگان
چکیده
One of the challenges often faced in forestry is the estimation of forest attributes for smaller areas of interest within a larger population. Small-area estimation (SAE) is a set of techniques well suited to estimation of forest attributes for small areas in which the existing sample size is small and auxiliary information is available. Selected SAE methods were compared for estimating a variety of forest attributes for small areas using ground data and light detection and ranging (LiDAR) derived auxiliary information. The small areas of interest consisted of delineated stands within a larger forested population. Four different estimation methods were compared for predicting forest density (number of trees/ha), quadratic mean diameter (cm), basal area (m2/ha), top height (m), and cubic stem volume (m3/ha). The precision and bias of the estimation methods (synthetic prediction (SP), multiple linear regression based composite prediction (CP), empirical best linear unbiased prediction (EBLUP) via Fay–Herriot models, and most similar neighbor (MSN) imputation) are documented. For the indirect estimators, MSN was superior to SP in terms of both precision and bias for all attributes. For the composite estimators, EBLUP was generally superior to direct estimation (DE) and CP, with the exception of forest density. Résumé : Un des défis souvent rencontrés en foresterie est l’estimation des attributs forestiers pour de plus petites zones d’intérêt au sein d’une population plus large qu’est la forêt. Les méthodes d’estimation pour petites zones sont des techniques bien adaptées à l’estimation d’attributs forestiers sur de petites superficies pour lesquelles la taille de l’échantillon existant est faible et l’information auxiliaire est disponible. Quatre méthodes d’estimation pour petites zones ont été sélectionnées et comparées pour estimer une variété d’attributs forestiers sur de petites superficies à l’aide de données terrain et de données auxiliaires dérivées du lidar. Les deux premières méthodes étaient indirectes (la prédiction synthétique (PS) et l’imputation par les plus proches voisins (PPV)); les deux autres étaient composites (la meilleure prédiction empirique linéaire sans biais (PSB) basée sur les modèles de Fay–Herriot et la prédiction composite basée sur la régression linéaire multiple (PC)). Les petites superficies d’intérêt étaient représentées par les peuplements délimités dans une forêt. Les attributs forestiers qui ont été comparés étaient la densité de la forêt (nombre de tiges/ha), le diamètre moyen quadratique (cm), la surface terrière (m2/ha), la hauteur maximale (m) et le volume des tiges (m3/ha). La précision et le biais des quatre méthodes d’estimation sont documentés. Dans le cas des estimateurs indirects, l’imputation par les PPV était supérieure à la PS en termes de précision et de biais pour tous les attributs. Dans le cas des estimateurs composites, la PSB était généralement supérieure à l’estimation direct et la PC, excepté pour la densité de la forêt. [Traduit par la Rédaction]
منابع مشابه
Comparison of Small Area Estimation Methods for Estimating Unemployment Rate
Extended Abstract. In recent years, needs for small area estimations have been greatly increased for large surveys particularly household surveys in Sta­ tistical Centre of Iran (SCI), because of the costs and respondent burden. The lack of suitable auxiliary variables between two decennial housing and popula­ tion census is a challenge for SCI in using these methods. In general, the...
متن کاملSpatial variability and estimation of tree attributes in a plantation forest in the Caspian region of Iran using geostatistical analysis
This research was conducted to investigate spatial variability and estimate tree attributes in a plantation forest in the Caspian region of Iran using geostatistical analysis. Sampling was performed based on a 50m?125m systematic grid in a maple stand (Acer velutinum Boiss) 18 years of age using circular samples of 200m2 area. Totally, 96 sample plots were measured in 63 hectares and 14.25 he...
متن کاملEstimation of stand volume by fusing low laser-sampling density LiDAR data with QuickBird panchromatic imagery in closed-canopy Japanese cedar (Cryptomeria japonica) plantations
This study investigated the potential of fusing low laser-sampling density LiDAR data with QuickBird panchromatic imagery for estimating stand volumes. The study area was in closed-canopy, mountainous Japanese cedar (Cryptomeria japonica) plantations in Japan. Stand volume in the area ranged from 250.5 to 913.1 m/ha and terrain was undulating with an elevation ranging from 135 to 391 m above se...
متن کاملSmall Area Estimation of the Mean of Household\'s Income in Selected Provinces of Iran with Hierarchical Bayes Approach
Extended Abstract. Small area estimation has received a lot of attention in recent years due to necessity demand for reliable small area statistics. Direct estimator may not provide adequate precision, because sample size in small areas is seldom large enough. Hence, by employing models that can use auxiliary information and area effects in descriptions, one can increase the precision of direct...
متن کاملTHE COMPARISON OF TWO METHOD NONPARAMETRIC APPROACH ON SMALL AREA ESTIMATION (CASE: APPROACH WITH KERNEL METHODS AND LOCAL POLYNOMIAL REGRESSION)
Small Area estimation is a technique used to estimate parameters of subpopulations with small sample sizes. Small area estimation is needed in obtaining information on a small area, such as sub-district or village. Generally, in some cases, small area estimation uses parametric modeling. But in fact, a lot of models have no linear relationship between the small area average and the covariat...
متن کاملذخیره در منابع من
با ذخیره ی این منبع در منابع من، دسترسی به آن را برای استفاده های بعدی آسان تر کنید
عنوان ژورنال:
دوره شماره
صفحات -
تاریخ انتشار 2011