Tip of the Week

Thursday, November 6th, 2014

Tip: Remote Sensing Utilities

GDAL

Command Line Utilities

  • Useful for almost all preprocessing
    • Metadata: gdalinfo
    • Reproject: gdalwarp
    • Format conversion: gdal_translate
    • Stack / mosaic: gdal_merge.py
    • Vector to Raster: gdal_rasterize
  • module load gdal/1.10.0
  • http://www.gdal.org/gdal_utilities.html

API

Write your own code? Use GDAL API:

  • Python
    • module load gdal/1.10.0 (loads python/2.7.5)
    • from osgeo import gdal
  • R (via rgdal and raster packages)
    • module load R_earth/3.1.0
    • library(raster)
  • C/C++:
    • module load gdal/1.10.0
    • #include "gdal.h"
    • GDAL_LIB and GDAL_INC paths

OGR

Command Line Utilities

  • ogrinfo
    • Basic information: ogrinfo -so -al vector.shp
    • Simple queries: ogrinfo -al -where "FEATURE > 5" vector.shp
    • SQL for more advanced queries (-sql)
    • Combine with grep, wc, etc... to quickly summarize vector data
  • ogr2ogr
    • Format conversion: ogr2ogr -f KML vector.kml vector.shp
    • Reproject: ogr2ogr -t_srs "EPSG:4326" vector_wgs84.shp vector.shp
    • Merge: ogr2ogr -append combined.shp part2.shp
  • http://www.gdal.org/ogr_utilities.html

QGIS

QGIS v2.0

  • On par with ArcGIS in some regards, better than in most.
  • Built on GDAL/OGR
  • QGIS vs ENVI
  • Plugin architecture
  • module load qgis/2.0.1

Landsat Data

Download

Bulk Download Application for L1 products:

  • module load bda/1.1.2
  • Use qsh session for big downloads -- very CPU intensive!
  • Updating is grief inducing... ask IT if update is required

Preprocess

Parallel SR / Fmask for Landsat 4, 5, 7, and 8 in one command:

  • module load batch_landsat/v4
  • landsatPrepSubmit.sh -h
  • Surface reflectance only suggested for Landsat 8 (as of Nov 6th, 2014)
    • LEDAPS for Landsat 4/5/7 from ESPA
    • Landsat 8 version still in development
    • We have ancillary data until 2014 DOY 282
  • See tutorial from May 2014

Visualization

Timeseries interactive visualization in QGIS:

  • "Stable" - module load CCDCTools/latest
  • "Development" - module load CCDCTools/_beta

Version control system (VCS)

Git

Reproducibility

IPython Notebook

Use Python? Want to share or explain your work, or just save and document your explorations?

KnitR

Installed as part of RStudio!