team-atlas

Team Atlas backlog

Backlog and team information for Team Atlas. The backlog is kept in the issues of this repo. Furher information on projects can be found in the wiki.

For an overview of what we are doing this sprint(a 2 week period) see the Sprint Board

The Project maps below serve as a high level overview of all the goals, deadlines, and stories (todo items) for a project. Click on each map for a large version, with clickable stories, goals, and elements.

Contributions to external projects, packages, tools, etc. are listed here

eEcolidar

Phenology Generalization

• Generalization of eScience tools
• Notebooks for education
• Comparison between Dask and Spark
• Publish previous work

Ice Shelves

This project aims at identifying and quantifying damage (sheaar distortions leading to crevasses etc.) in antarctic ice shelves from satellite remote sensing data (optical/NIR imaging as well as (In)SAR), with the goal of providing a workflow and platform to monitor ice shelves and their evolution over the entire continent on an on-going basis.

Global Vegetation

The goal of this project is to use direct (backscatter) and indirect (slope and curvature) ASCAT observations to monitor the vegetation water processes and detect the anomalies.

According to the proposal and previous work from Manuel and Xu, we are going to achieve this by three steps:

• Step 1: Develop a surrogate model based on ML to map land-surface model parameters (input) to ASCAT obervations (output).
• Step 2: Study the sensitivity of ASCAT obervations to certain parameters in land-surface model. This will be done by adding very small perturbations from an assumed distribution to the model states, and quantify the sensitivity of the ML-derived model. This is to ensure that the surrogate model is robust.
• Step 3: Data assimilation. Use the surrogate model as a measurement operator in an assimilation framework to assimilate the ASCAT observables into the land-surface model. An example of the assimilation framework is LDAS-Monde assimilation system, which only works for soil moisture. Current plan is open to other framework or implement their own assimilation system.

Eratosthenes

The Eratosthenes project aims at using the shadows that mountain peaks cast on glaciers to estimate the glaciers' elevation change. Repeat high-resolution satellite images will be used to determine the displacement of the shadow edge, which will then be translated into a vertical elevation difference.

Dark Generators

First three goals:

• Simulators: CMB, Lensing, LHC
• Parallel computation
• Network optimization