Activity Number:
|
433
- SPEED: Applications of Advanced Statistical Techniques in Complex Survey Data Analysis: Small Area Estimation, Propensity Scores, Multilevel Models, and More
|
Type:
|
Contributed
|
Date/Time:
|
Tuesday, July 31, 2018 : 2:00 PM to 2:45 PM
|
Sponsor:
|
Survey Research Methods Section
|
Abstract #332860
|
|
Title:
|
Empirical Bayes Small Area Prediction of Sheet and Rill Erosion Using a Zero-Inflated Lognormal Model
|
Author(s):
|
Xiaodan Lyu* and Emily Berg and Heike Hofmann
|
Companies:
|
Iowa State Univ and Iowa State University and Iowa State University
|
Keywords:
|
small area prediction;
empirical bayes;
zero inflated lognormal;
rainfall erosion loss;
RUSLE2;
CEAP
|
Abstract:
|
In the Conservation Effects Assessment Project (CEAP), some of the variables are skewed right and have zeros. We proposed an empirical Bayes (EB) estimator for the Zero-inflated Log-normal model. We are also able to approximate the variance of the estimator by estimating the conditional variance. A Monte Carlo simulation is conducted to show the empirical properties of the EB predictor and its efficiency gain over the Plugin and Zero-ignored Minimum Mean Squared Error estimator. We also applied the method to the CEAP data of South Dakota, where there are about 15% zeros among the observed RUSLE2. The proposed method is applied to get the predicted population mean of rainfall-erosion losses from cropland at the county level for South Dakota. For data analysis, we overlaid soil map shapefile and Cropland Data Layer raster in R to obtain a list of cropland map units for South Dakota, which is our target population for small area prediction. A Shiny app visualizing the overlaying procedure is developed.
|
Authors who are presenting talks have a * after their name.