v Abstract: The Enhanced Forest Inventory and Analysis program of the U.S. Forest Service has established a nation- r l wide array of permanent field plots, each representing approximately 2400 ha. Each plot has been assigned to one of five interpenetrating, nonoverlapping panels, with one panel selected for measurement on a rotating basis each year. As a with most large surveys, observations for some plots are not obtained. Several techniques for dealing with missing ob- servations were compared with respect to the bias and precision of stratified estimates of mean volume per unit area using data from the State of Indiana in the north-central region of the United States for illustrative purposes. For the proportion of missing plots ranging from 1% to lo%, acceptable results were obtained for techniques that both ignored and replaced missing plot observations, but only when the techniques accounted for differences between plots in private and public ownerships. R&sumi: : Le programme ame'liork d'inventaire forestier et d'analyse du Service forestier des ~tats-~nis a permis d'e'tablir un dispositif national de placettes permanentes. Repre'sentant approximativement 2400 ha, chaque placette a e'te' assigne'e B l'un des cinq groupes qui s'interpdn5trent sans chevauchement. Chaque groupe est se'lectionnk B tour de rcile pour &re mesure' sur une base annuelle. Gomme dans tous les inventaires B grande kchelle, les observations de cer- taines placettes manquent. Pour resoudre ce problkme d'observations manquantes, plusieurs techniques ont e'te' compa- r6es quant au biais et ri la pre'cision des estimations stratifie'es du volume moyen par unite' de surface, en utilisant B titre d'exemple les donne'es de 1'~tat de 1'Indiana situk dans le centre nord des ~tats-~nis. Lorsque la proportion des placettes rnanquantes se situe entre I et 10 %, des resultats acceptables ont e'te' obtenus tant par les techniques qui ignorent que celles qui remplacent les observations des placettes manquantes, mais seulement lorsque ces techniques tiennent compte des diffe'rences entre les placettes sur les terres publiques et celles sur les terres prive'es. (Traduit par la Re'daction)
[1]
S. Lemeshow,et al.
Nonresponse in Sample Surveys
,
2004
.
[2]
Mark H. Hansen,et al.
Using a land cover classification based on satellite imagery to improve the precision of forest inventory area estimates
,
2002
.
[3]
Limin Yang,et al.
COMPLETION OF THE 1990S NATIONAL LAND COVER DATA SET FOR THE CONTERMINOUS UNITED STATES FROM LANDSAT THEMATIC MAPPER DATA AND ANCILLARY DATA SOURCES
,
2001
.
[4]
R. McRoberts.
Joint Annual Forest Inventory and Monitoring System: The North Central Perspective
,
1999
.
[5]
Gregory A. Reams,et al.
The Southern Annual Forest Inventory System
,
1999
.
[6]
P. Deusen,et al.
Annual forest inventory statistical concepts with emphasis on multiple imputation
,
1997
.
[7]
Mark H. Hansen,et al.
Cubic and Board Foot Volume Models for the Central States
,
1991
.
[8]
D. Rubin,et al.
Multiple Imputation for Interval Estimation from Simple Random Samples with Ignorable Nonresponse
,
1986
.
[9]
Innis G. Sande,et al.
Imputation in Surveys: Coping with Reality
,
1982
.
[10]
J. Gove,et al.
Forest inventory
,
2001
.
[11]
M. Hansen,et al.
Indiana's Forests in 1998.
,
2000
.
[12]
G. Reams,et al.
The use of multiple imputation in the Southern Annual Forest Inventory System
,
2000
.
[13]
Mark H. Hansen,et al.
Using classified Landsat Thematic Mapper data for stratification in a statewide forest inventory
,
2000
.
[14]
A. Winsor.
Sampling techniques.
,
2000,
Nursing times.
[15]
M. D. Nelson,et al.
The hexagon/panel system for selecting FIA plots under an annual inventory
,
2000
.
[16]
Smith.
Factors and equations to estimate forest biomass in the North Central region. Forest Service research paper (Final)
,
1985
.
[17]
Jerold T. Hahn,et al.
Tree volume and biomass equations for the Lake States.
,
1984
.
[18]
Fritz Loetsch,et al.
Forest inventory. Volume I. Statistics of forest inventory and information from aerial photographs.
,
1964
.