Presenter(s):
Debasish Mishra, Texas A&M University College Station
Leah Kocian, Texas A&M University
Frank Anyoka Adekilae, Louisiana State University
Vinit Sehgal, Louisiana State University

Primary Convener:
Vinit Sehgal, Louisiana State University

Analysis of large-scale geospatial data (regional- to global-scale at high spatial and temporal resolution) can be computationally expensive and time-consuming, especially when working with multiple formats and sources of data. R provides a powerful computational alternative to popular Geographic Information System (GIS) software to organize, analyze, and visualize geospatial data. R enjoys a vast collection of open-source libraries for GIS-type operations and proven statistical analysis and data visualization capabilities. Taking examples from global satellite data in gridded/ raster format, we will demonstrate several geospatial operations like projections, resampling, spatial extraction, cropping, masking, etc., using rasters, shapefiles, and spatial data frames. For a seamless analysis across different datatypes and platforms, conversion from/to different data formats like data frames, matrices, raster, and structured data like NetCDF will be discussed. Advanced topics will include working with multilayer raster/ raster time series, layer-wise operations on multilayer raster, and cell-wise operations on raster time series by implementing user-defined functions. Out-of-the-box multicore parallel application of user-defined functions for block-, layer-, and cell-wise parallel operations will be demonstrated on large-scale datasets.

• Instruction type: Hands-on.


• Course level: Advanced


• Course requirements: Computer with installed and working R and RStudio. Working experience in R is required.