geohackweek / ghw2018_landlab

The repository contains curriculum developed for geo-spatio-temporal analysis using open source research software

Project Title: Advanced Geo-spatial-temporal research software curriculum: Developing Landlab Modules to teach open source methods

[Paragraph describing the need and benefits to open source research software education in earth sciences]

Application Example and Motivation [Landlab group on HydroShare using CI to train new students on a science domain specidfic modeling framework - How would we do that in other domains?

Sample Application and Rubric [Rose - Georgia Tech example - EdX; IDE for Python Enthought; Decribe rubric as compared to online tutorials e.g. ESRI certificated, CUAHSI Virtual Classroom]

Specific Questions (breaking up of the project task) The project contains two main parts: designing the framework for new modules and development of individual modules

Landlab Background Links:

Background on the model grid data model used in Landlab

Teach yourself Landlab

1. The Framework

The repository contains curriculum developed for geo-spatio-temporal analysis using open source research software.

Start to Finish Tutorial building for Geo-spatio-temporal education and research

We are developing a framework for research software development that applies to a range of examples usable in the classroom.

Common Workflow Tools

Github - describe why, how

HydroShare This JupyterHub is one way to launch

Landlab

JupyterNotebooks

UserExperience

Formal Publication

Distribution Channel Strategies

Component based software design

Contribute to an Open Source Project

2. Modules

Each modules uses common workflow tools that we share developing, using and teaching each other. Each module uses existing open source code and explores advancing this code with at least one new component.

Steven's Module

This project is a modification of a Landlab tutorial illustrating use of Python and Jupyter for landslide hazard estimation. It uses factors feeding into the infinite slope factor of safety equation, which predicts the ratio of stabilizing to destabilizing forces on a hillslope plane. The properties are assumed to represent an infinte plane, neglecting the boundary conditions around the landslide location. The Jupyter notebook, which is hosted and runs on Hydroshare, is a Monte Carlo simulation for the Fisher Creek watershed in the North Cascades; the exercise is adapted from work by Stauch et al. (https://www.hydroshare.org/resource/a5b52c0e1493401a815f4e77b09d352b/).

The goal is/was to explore possible teaching resources for graduate level students in the Department of Earth and Space Sciences, many of whom conduct research on landslide hazards in western Washington. The open source geospatial analysis and modeling tools afforded through Python (and ease of their teaching and use through Jupyter notebooks) is extremely valuable.

Link to the tutorial github repository: https://github.com/swalt826/Landlab---Landslide-Hazards

Rose's Module

This tutorial is an introduction on how to use spectrogram cross-correlation to detect blue whale calls. This tutorial is meant primarily for individuals with at least an introductory understanding of Python to familiarize themselves with the one of the processes used in acoustic signal detection.

Workflow tools in the tutorial:

-Python: Numpy, Scipy, and Matplotlib libraries

-Jupyter notebook

-Github

Key tutorial outcomes:

-Experimentation with key scientific python libraries

-Introduction to timeseries analysis

-Familiarization with spectrograms

-Understand basic reasoning behind signal detection

Link to the tutorial github repository: https://github.com/rosehilmo/whale-detection

Helen Xarray Module

Tutorial link:https://github.com/hydrogeohc/Xarray

Extra file in and out function by using Xarray.

Load time series Geotiff raster files

Create Xarray data structure

Save it as netCDF files format

Export the point of interest for this netCDF file

Christina's Modules

Background on the model grid data model used in Landlab

Teach yourself Landlab

Test the Modules from this HydroShare Resource (requires HydroShare UserID to use the cloud compute resources to run interactive models from CUAHSI JupyterHub). Lowering the barriers to Computational Modeling of the Earth Surface

Sign in/Sign up at www.hydroshare.org. Go to Collaborate. Search for the Landlab Group. Join. Go to Resources.

Click on the resource above. Use blue upper right button to 'OPEN WITH'. Select JupyterHub.

Execute Welcome to connect to HydroShare. Click on Jupyter Notebook of interest.

ghw2018_landlab's Issues

Updates from Day 3 for Day 4

• Run through/modify Landlab landslide tutorial

• Add visualizations(?) to landslide tutorial

• Explore alternate sites in NOCA for landslides

Start to Finish Day 1-2

• Practice Github

• Run a Landlab Module .ipynb from HydroShare

• Run a .py from the cloud

• Make a Team strategy

• Think about distribution option #1 Landlab Tutorials repo

• Learn a new package; Landlab

Updates for Landslide Fire tutorial

• Get Node ID shapefile for all of NOCA from Ronda

• Add text to explain how to make a new selection, publish as a new HydroShare resource, change section in tutorial for the new subset

Geospatial analysis undergrad project: watershed selection for Landlab modeling in Python

Tasks

• Open North Cascades model extent in Geo Views

• Open a watershed shapefile in Geo Views; or open all HUC watersheds and select one.

• Select NOCA Landlab GridCodeID that intersect the watershed boundary and 30m grid cell centroids.

• Create new node point subset shapefile

• Publish on HydroShare

• Update model tutorial to use new subset

For Friday

• Make folders and push/pull data to this repo for copies to be saved with eScience

• Work on presentation

• Make references page

• Update Helpful GIthub link

• Make distribution strategy