Online Program

Return to main conference page
Thursday, May 30
Data Science Techologies
Practice and Applications
Data Science Applications E-Posters, II
Thu, May 30, 5:30 PM - 6:30 PM
Grand Ballroom Foyer

For-estimation: Post-stratification to increase efficiency of forest attribute estimates (306340)

Edwin Alvarado, Reed College 
Alex Lloyd-Damnjanovic, Reed College 
*Miranda Rintoul, Reed College 
Mai Toyohara, Reed College 

Keywords: poststratification, forestry, ecology, variance, estimation

The National Forest Inventory and Analysis (FIA) Program of the US Forest Service regularly estimates forest attributes such as biomass and trees per acre. These estimates are used in a wide variety of applications, such as policy formulation, scientific analysis, land management, and business plan development. The use of post-stratification in prediction of forest attributes can greatly improve the efficiency of these estimates. Current procedures used by the Interior West unit of the FIA involve stratifying by forest, non-forest, and water areas. This project aims to increase the efficiency of estimates by finding a better stratification scheme. We plan to propose different stratification schemes based on satellite image data, and compare their variance estimates to that of the current stratification scheme. This research will hopefully improve on the previous FIA stratification scheme by reducing the amount of variation in forest attribute estimation, which can in turn make for more useful and informative analyses.