Skip to content

Latest commit

 

History

History
68 lines (44 loc) · 6.32 KB

README.md

File metadata and controls

68 lines (44 loc) · 6.32 KB

GrIMPTools

This readme provides information on and links tools for working with data from the Greenland Ice Mapping Project, which are funded by the NASA MEaSUREs program and archived at NSIDC.

The Greenland Ice Mapping Project (GrIMP) has now archived several TB of image, velocity, elevation, and glacier terminus data at NSIDC, representing a far greater volume than many users are able to store on their laptops or other computers. Many studies require only a small subset of these data (e.g., for a particular glacier). As a result, the GrIMP project has developed several tools for accessing and working with the data remotely, eliminating the need for voluminous downloads in many studies.

The tools are all built from open source Python code that is called from Jupyter notebooks. The examples have been structured with default parameters to do much of what the typical user may wish to do, with little or no python knowledge.

These tools are distributed across several gitHub repositories, which can be reached by following the links below.

All tools can be run locally on users machines or run directly, with some limitations, on freely available cloud resources.

Binder

The notebooks in the repositories listed below can run locally on a users own machine or cloud resource, but users can also run them remotely free of charge as a binder app. Advantages to binder are:

  • The notebook runs in a well-tested, pre-configured environment,
  • No need to setup on your own machine.

There are some limitations to binder:

  • The binder instances are often limited in memory, which can easily be exceeded if too large a region is or too many products are selected.
  • The binder instance will time out after an ten minutes of inactivity, and all work will be lost. Keeping the notebook in the foreground while it runs should avoid a timeout. Products must be downloaded to be saved.

Installation on Local Machine

If not running in binder, the easiest and surest way to run these notebooks is use create a conda environment using the environment.yml file in the binder folder for this repository.

conda env create -f environment.yml
python -m ipykernel install --user --name=greenlandMapping

Once installation is complete:

conda activate greenlandMapping
jupyter lab

It the libraries don't load once the notebook starts to run, check that the greenlandMapping kernel is selected from the pulldown in the upper right corner of the screen.

The code in these notebooks requires code from three repositories: grimpfunc, nisardev, and grimpqgis. The latter is needed only for the qgisRremoteNotebook.ipynb notebook. These packages are installed automatically with the conda install described above. Otherwise, they can be pip installed directly from their respective github repos (pip >= v22.0.3).

GrIMPNoteBooks Binder (click icon to launch binder)

This repository contains notebooks to demonstrate tools for working with GrIMP data directly from the archive. Specifically the code can be used to:

  • Generate publication plots directly from the GrIMP products at NSIDC;
  • Create or download QGIS project and layer definition definitions files so that GrIMP products can be displayed directly from the NSIDC servers;
  • Subset data and download time series stacks directly to netCDF files.

GrIMPQGISProject Binder (click icon to launch binder)

This repository contains QGIS project (.gqs) and layer definition files (.qglr) that allow remote GrIMP products to be directly ingested and viewed in QGIS with no intermediate storage on the local machine.

These products require a NASA Earth Data Login, which requires a free registration.

The .qgs and .qlr files in this repository can be run directly by QGIS. Some minor setup is required to authenticate as described here NSIDCLoginNotebook.

Supporting Python Code

The notebooks described above using the following Python libraries:

Further Information

These tools were created by Ian Joughin and Scott Henderson at the University of Washington. For further information or to provide feedback contact: .