Cryovex Eureka Reseach Analysis Source Code

Evaluating Ice Surface Elevation Estimates using Airborne Radar Altimetry from the CryoVEX-Eureka 2014 Arctic Campaign

Author: Paul Donchenko

Special Thanks to:

​ Josh King (Environment and Climate Change Canada)

​ Richard Kelly (University of Waterloo)

Installation

Clone this repository to a directory of your choice.

Airborne L1B Data

Most of the data used in this analysis is located in the data folder. The L1B ASIRAS and ALS airborne data from the CryoVEx 2014 campaign is not provided.

To obtain this data, contact ESA and request the datasets listed below. Information about how to request data can be found at: https://earth.esa.int/web/guest/pi-community/apply-for-data/campaigns

or by its DOI: https://doi.org/10.5270/esa-aa4xtkn

By default, both files should be placed into data/l1b

ASIRAS

• File Name: AS3OA03_ASIWL1B040320140325T160941_20140325T164233_0001.DBL
• Date: 2014/03/25
• Product: L1B

ALS

• File Name: ALS_L1B_20140325T160930_164957
• Date: 2014/03/25
• Product: L1B

ECCC 2014 Ground Observations Data

A modified version of this dataset is included in the repository. The original can be found at the DOI: https://doi.org/10.5281/zenodo.823679

Anaconda 3 Environment

1. Install Anaconda 3 from https://repo.anaconda.com/archive/ or https://repo.continuum.io/archive/

   Preferably version 2019.10 for Windows 10, although future versions on other platforms are likely to work as well.

2. Create a new conda environment from the supplied cveureka.yml file by running conda env create -f "<path_to_project>/cveureka.yml" from the Anaconda prompt

   This should create a new environment in your Anaconda3/envs directory called cveureka. The environment will contain Python 3.7 and all the necessary packages.

   If you have issues creating the environment, try switching to conda version 4.7.12

PostgreSQL Database

1. Download PostgreSQL for your platform from https://www.postgresql.org/download/

   Alternatively you can use a remote PostgreSQL connection.

   This project was developed on version 10, so your version must be equal or greater. If you have issues with deprecated features, try using version 10.

2. Create a new target database if one doesn't exist

3. Install PostGIS in the target database. Steps will vary depending on platform https://postgis.net/install/

Configuration

1. Modify the Database section of config.ini to match the connection settings of your PostgreSQL database.

   default_schema should be changed to a schema specially prepared for this project. Using the public schema is not recommended since that is where PostGIS installs its functions, and moving the results tables after they are created can be tricky.

   default_geom_col should not be modified unless output tables will be inputs into a pipeline

2. Modify the data_dir variable in the Files section to match the location of the data folder which contains all of the input datasets needed to run the methods procedure. By default the folder is located inside the root repository directory.

   The individual dataset variables do not have to be modified and should sit inside the data folder.

Usage

The data is process in two parts: the method and the analysis. The method takes the raw input and produces PostgreSQL tables with the ice surface estimate and error results, which is equivalent to the manuscript Methods and Results section. The analysis reshapes parts of the results to create figures that are referenced in the Analysis and Discussion manuscript sections.

Method

The src/method.py script is responsible for taking the input data and producing output tables in the PostgreSQL database which have ice surface estimates and their associated error.

To run the method procedure, activate the cveureka conda environment, and then run the method.py script as module src.method with the activated python environment. The path to the config.ini should be the first and only argument to the script. The script must be a run as a module due to the use of relative imports.

To run in Windows, use the following commands with the repository root folder as the working directory:

conda activate
python -m src.example "config.ini"

A batch file method.bat is provided with default configuration for running in Windows.

Analysis

The src/cve_analysis directory contains R scripts that connect to the PostgreSQL database, consume the results and produce the analysis figures:

• config.r stores processing constants and reads configurations from config.ini in the project root
• tools.r contains helper functions for reshape and analyzing the results
• scripts that begin with plot_ generate the manuscript plots into the plots directory in the project root

None of the scripts need to modified to produce the default results. If config.ini cannot be found the process will ask for its location.

It is recommended to use RStudio to run the scripts as it should retrieve and install the necessary packages automatically.

Run the plot_err_all.r script to produce all plots.

Documentation

A description of the L1B binary format used to store the airborne ALS and ASIRAS data is available in docs/cryovex_airborne_data_description.pdf

Descriptions for output tables and columns can be found in docs/table_info.md