You are here:
An application of spatio-temporal modeling to finite population abundance prediction
Citation:
Higham, M., M. Dumelle, C. Hammond, J. Ver Hoef, AND J. Wells. An application of spatio-temporal modeling to finite population abundance prediction. JOURNAL OF AGRICULTURAL, BIOLOGICAL, AND ENVIRONMENTAL STATISTICS. American Statistical Association, Alexandria, VA, , s13253-023-00565-y, (2023). https://doi.org/10.1007/s13253-023-00565-y
Impact/Purpose:
This work shows how incorporating both spatial and temporal information into analysis can benefit environmental monitoring programs, yielding more precise estimates of population parameters. We provide relevant statistical background and point readers to software they can use to implement the methodology themselves.
Description:
Spatio-temporal models can be used to analyze data collected atvarious spatial locations throughout multiple time points. However, even witha finite number of spatial locations, there may be a lack of resources to sampleevery spatial location at every time point. We develop a spatio-temporalfinite-population block kriging (ST-FPBK) method to predict a quantity ofinterest, such as a mean or total, across a finite number of spatial locations.This ST-FPBK predictor incorporates an appropriate variance reduction forsampling from a finite population. Through an application to moose surveysin the east-central region of Alaska, we show that the predictor has a substantiallysmaller standard error compared to a predictor from the purely spatialmodel that is currently used to analyze moose surveys in the region. We alsoshow how the model can be used to forecast a prediction for abundance ina time point for which spatial locations have not yet been surveyed. A separatesimulation study shows that the spatio-temporal predictor is unbiased and that prediction intervals from the ST-FPBK predictor attain appropriatecoverage. For ecological monitoring surveys completed with some regularitythrough time, use of ST-FPBK could improve precision. We also give an Rpackage that ecologists and resource managers could use to incorporate datafrom past surveys in predicting a quantity from a current survey.
URLs/Downloads:
DOI: An application of spatio-temporal modeling to finite population abundance prediction