Andrew Finley

Departments of Forestry and Geography

Geospatial Lab - Natural Resources Building - Michigan State University

image

Overview

The Geospatial Lab is located in the Natural Resources Building at Michigan State University.

Most of the research conducted in the lab relies on our small, but growing, Linux cluster and several large heterogeneous (Xeon/Phi co-processor) machines, see, e.g., HPC notes.

The lab is currently supported by several ongoing NSF, NASA, and USDA Forest Service projects.

Current lab members

Malcolm Itter

I recently completed a MS in Statistics from Oregon state University and am currently working on a PhD in Forestry. My research interests are in developing statistical models to better understand forest change due to succession, natural disturbance, management, and climate change. The central theme of my research is the use of rigorous statistical methods to answer important questions in forest ecology.

Megan Kress

I am currently working as a research assistant and will start a MS in Forestry in Fall 2015. I graduated from the University of Wisconsin with a BS in Mathematics and minor in Computer Science. I am interested in linear programming to optimize environmental resource allocation problems. With a background in education, I am also interested in promoting ecoliteracy by enhancing the connections between ecology and mathematics in STEM education.

Neil Ver Planck

I am currently working towards a PhD in Forestry. My research interests lie in the modeling of forest attributes, specifically improvements to biomass estimation and the associated uncertainty.

Abolfazl Safikhani

I am a doctoral candidate in Statistics. My main research interests lie in spatial statistics, space-time modeling, and estimation and prediction in big spatial datasets. More specifically, these interests include developing nonstationary spatial models, covariance estimation in big spatial data, and hierarchical multivariate Gaussian spatial models.

Yuzhen Zhou

I am a PhD candidate in statistics. My research focus is on theoretical aspects of multivariate spatial models, Bayesian hierarchical modeling for high dimensional space and space-time data, and extreme theory and fixed domain asymptotics of multivariate Gaussian random fields.

Past lab members

Chad Babcock

Graduate student. Completed MS in Statistics and Geography, 2014. Research focused on developing spatial-temporal models to estimate present and future forest characteristics.

2014, off to do a PhD at University of Washington. Find his current site here.

Paul Delamater

Graduate student. Completed PhD in Geography, 2013. Research focused on spatial epidemiology, access to healthcare services, and modeling spatial and temporal trends in healthcare service utilization.

2014, off to a Assistant Professor position at George Mason. Find his current site here.

Clara DeYoung

Undergraduate student. Completed BS in Forestry, 2011. Research interests included forest growth trends under changing environments.

2011, off to do a MS at Virginia Tech.

Santonu Goswami

Postdoctoral researcher. PhD in Environmental Science and Engineering in 2011 at the University of Texas at El Paso.

2012, off to a postdoctoral position at Oak Ridge National Laboratory.

Virpi Junttila

Postdoctoral researcher. PhD in applied mathematics in 2011 at the Lappeenranta University of Technology, Finland. Postdoctoral research focused on remote sensing data based forest inventory, especially Bayesian methods to improve the precision of inventory parameter estimates.

2013, off to a research faculty position at Lappeenranta University of Technology.

Jason Matney

Graduate student. Completed MS in Geography, 2014. Research focused on quantifying uncertainty in permafrost predictions across interior Alaska.

2014, off to do a PhD at University of NC State. here.

Francesc Montane

Postdoctoral researcher. PhD in 2010 at the University of Barcelona, Spain. Postdoctoral research focused on models for quantifying carbon cycling, vegetation dynamics and global change.

2012, off to a postdoctoral position at University of Oklahoma.

Brian Walters

Graduate Student. Completed MS in Geography, 2010. Research focused on developing methodology to create large area forest attribute maps using Forest Inventory and Analysis (FIA).

2010, off to a permanent position as an analyst with US Forest Service FIA ... dreams really do come true.

Zhen Zhang

Graduate Student. Completed a MS and PhD in statistics, 2014. Research interest focused on Bayesian computing and spatial modeling, and statistical pattern recognition.

2014, off to a postdoctoral position at Department of Statistics, University of Chicago.

Huirong Zhu

Graduate Student. Completed a MS in Statistics and Geography, 2011. Research focused on quantifying changing forest attributes in human-natural systems. General interests include geostatistics, Bayesian methodology, spatial hierarchical modeling, and statistical computing.

2011, off to do a PhD in Statistics at University of Texas Health Science Center.

Lab resources

HOWTOs using R, GDAL, and other open source tools

Downloading and processing National Oceanic and Atmospheric Administration (NOAA) hourly weather station data.
Details and Code

Downloading and processing The North American Regional Climate Change Assessment Program (NARCCAP) data.
Details and Code

Downloading and processing Environmental Protection Agency (EPA) pollution monitoring data.
Details and Code

Computing notes

Some considerations for fitting geostatistical models to large datasets using the C/C++/Fortran libraries on shared memory multiprocessing systems.
Details

Dirk Eddelbuettel provides an excellent description and benchmark analysis for several of the same BLAS/LAPACK implementations available on our sprout and ficus workstations.
Details

Coding examples

Statistical computing examples for shared memory multiprocessing systems.
Details

R Tutorials

Andrew Robinson has updated his excellent IcebreakeR tutorial.
Details