Title	Author	Date
Climate Data Demo	John Fay	Fall 2018

Climate Data Demo

Objectives

Explore how Python can read, manipulate, and visualize multi-dimensional climate data.

0 - Set up

• Getting Python
  • Installing miniconda
  • Using Duke's Jupyter containers
  • Using the ArcGIS Python installation
• Installing packages:
  • pip install --upgrade pip
  • pip install netCDF4
  • conda install xarray
• Using Jupyter

  • Basic navigation

  • Importing the Python notebooks

    git clone https://github.com/ENV859/ClimateData

1 - Fetching the data:

• 1-FetchData.ipynb notebook

  • Pulls data to local workspace using Python's urllib package.

2 - Examining the data with NetCDF4, NumPy, and Pandas:

• 2-Numpy-and-Pandas.ipynb notebook
  1. Import the .nc file into a netCDF4 dataset object.
  2. Exploring the netCDF dataset:
     1. Dimensions
     2. Variables
     3. Attributes (global and variable)
  3. Working with the data
     1. ...as netCDF variable objects...
     2. NetCDF variable shapes
     3. Extracting single data points from NetCDF variables
  4. Using NumPy with netCDF variables
     1. Converting netCDF variables to Numpy [masked] arrays
     2. Working with NumPy arrays
     3. Subsetting & visualizing our data
        1. Time series for a single location
        2. Location matrix for a single time
        3. Plotting
  5. Using Pandas with netCDF variables
     1. Reducing arrays and converting to dataframes
     2. Adding indexes and column names
     3. Subsetting data
  6. Xarrays...

3 - Examining the data with Xarrays

• 3-xarrays.ipynb notebook.
  1. Reading data into an xarray dataset object
  2. Extracting variables into xarray data array objects
  3. Working with xarray data arrays
     1. Extracting data for one point
     2. Plotting a time series for one location
     3. Mapping precipitation for one time point
     4. Creating spatial subsets
  4. Calculating and plotting summer averages