1. Introduction

1.1. Overview

Geo2Grid is a set of command line tools for extracting data from earth-observing satellite instrument files, remapping it to uniform grids if needed, and writing that gridded data to a new file format. It provides an easy way to create high quality projected images. Geo2Grid was created by scientists and software developers at the SSEC. It is distributed as part of the CSPP Geo project for processing of data received via direct broadcast antennas. Although Geo2Grid was created to serve the direct broadcast community, it can be used on most archived data files.

The features provided by Geo2Grid are accessible via bash scripts and binary command line tools. This is meant to give scientists an easy way to use and access features that typically involve complicated programming interfaces. Linux terminal commands included in these instructions assume the bash shell is used.

GitHub Repository

CSPP Geo2Grid Forum

1.2. Software Design

digraph glue_flow {
    rankdir = LR;
    node [shape = rectangle, fillcolor="#C3DCE7:white", gradientangle="90.0", style="filled"];
    "Reader" -> "Remapper";
    "Remapper" -> "Writer";
    "Remapper" -> "Compositors" [style=dashed];
    "Compositors" -> "Writer" [style=dashed];
}

Geo2Grid has a modular design operating on the idea of satellite “products” or “datasets”; data observed by a satellite instrument. These products can be any type of raster data, such as temperatures, reflectances, radiances, or any other value that may be recorded by or calculated from an instrument. There are 4 main steps or components involved when working with these products in Geo2Grid: reading, writing, compositing, and remapping. Geo2Grid makes it possible to access and configure these steps with a simple bash script called geo2grid.sh and other helper scripts. More information on accessing Geo2Grid’s features and running its scripts can be found in the Geo2Grid Basics section or the examples following each reader section. Note that although an example may be written for a specific reader the same operations can be applied to all readers unless mentioned otherwise.

For more low-level information on the design and responsibility of each component see the Software Design Overview Appendix.

In Geo2Grid a majority of the functionality is provided by the open source SatPy library created by the PyTroll group. More information on SatPy and the capabilities it provides to python users can be found in the SatPy documentation.

1.3. What’s New?

Included in this release:

  • New Geo2Grid Package!
  • ABI L1B (abi_l1b) reader added
  • AHI HSD (ahi_hsd) reader added
  • AHI HRIT/HimawariCast (ahi_hrit) reader added
  • Geotiff (geotiff) writer added
  • Multi-threaded (multiple worker) processing
  • Sharpened rayleigh-corrected full-resolution true and natural color RGBs
  • Command line Lat/Lon defined subsets
  • User defined grid capability
  • MIN/MAX native resampling possible

For more details on what’s new in this version and past versions see the Release Notes in the github repository.

1.4. System Requirements

For minimal processing requirements (i.e. not realtime) the following system specifications are required:

  • Intel or AMD CPU with 64-bit instruction support (2+ cores - 2.4GHz)
  • 16 GB RAM (minimum)
  • CentOS 6 64-bit Linux (or other compatible 64-bit Linux distribution)
  • 5 GB disk space (minimum)

For a more demanding processing load, like realtime generation of all GOES-16 ABI channels, true color, and natural color RGB images at full resolution, a system should have at least:

  • Intel or AMD CPU with 64-bit instruction support (20+ cores - 2.4GHz)
  • 64 GB RAM (minimum)
  • CentOS 6 64-bit Linux (or other compatible 64-bit Linux distribution)
  • 1 TB disk space (minimum for ~1 week of images, does not include long-term storage)

Local storage (i.e. not network file systems) are preferred to limit any effect that network congestion may have. If additional satellites are included in the processing requirements then the above system requirements will need to be adjusted accordingly.

1.5. Execution Times

The following table provides execution time averages for creating all default GeoTIFF images at full spatial resolution for the given instrument and sector. Eight computer threads were used. The times are provided for the higher end system defined above. Execution times decrease when fewer bands and smaller regions are processed.

Table of Execution Times for Creating GeoTIFF default Images (All bands plus true and natural color images)

Instrument Full Disk Sector CONUS Sector 1000x1000 pixel subset
GOES ABI 6m50s 1m04s 31s
AHI HSD 13m03s N/A 40s
AHI HimawariCast 1m32s N/A 29s

1.6. License and Disclaimer

Original scripts and automation included as part of this package are distributed under the GNU GENERAL PUBLIC LICENSE agreement version 3. Software included as part of this software package are copyrighted and licensed by their respective organizations, and distributed consistent with their licensing terms.

The University of Wisconsin-Madison Space Science and Engineering Center (SSEC) makes no warranty of any kind with regard to the CSPP Geo software or any accompanying documentation, including but not limited to the implied warranties of merchantability and fitness for a particular purpose. SSEC does not indemnify any infringement of copyright, patent, or trademark through the use or modification of this software.

There is no expressed or implied warranty made to anyone as to the suitability of this software for any purpose. All risk of use is assumed by the user. Users agree not to hold SSEC, the University of Wisconsin-Madison, or any of its employees or assigns liable for any consequences resulting from the use of the CSPP Geo software.