Session 1 - NASA Updates


HQ Perspectives - Katherine's presentation:

  • HQ shared a new objective for user community, data archive centers and ESDIS to envision ways to support the growth and sustainability of the Earth Science to Action initiative. This initiative aims to support openness through programs and tools like UNBOUND, Veda.
  • We were challenged to think of existing technologies that should be leveraged now and of new ones that could be developed to sustain growth necessary to support future research and data goals
  • UWG mentioned that pathfinders are a good way to capture users
  • Katherine mentioned that pathfinder postcards are shared at events
  • UWG mentioned NOAA has a program similar to the Earth Science to Action application model, which sparked a conversation about what NASA's user focus should be (power user, mid-level, beginner, all) and how to leverage relationships with other organizations to enable us to provide users with a full range of services from all organizations. These relationships were noted as necessary to span this diverse user base, to maintain the overall mission of NASA (e.g., sensor quality and advancing the state of the science) while also expanding Earth observations, such as through products that provide global coverage by interpolation of merged satellite observations (see notes below OBB Perspectives for further context that developed over the course of the meeting)


  • Challenges (shared by UWG):
    • size of data is so large such that it is more desirable to work with it/process in the cloud rather than download it which is time consuming and quickly fills storage
    • data format (netCDF) is a barrier. Users ultimately want to have the option to convert to a csv or html file to work with
    • data quality concerns arise from no clear links from dataset to documented processes for product generation
    • Data management plan is meant to provide that information and clarity to the user


OBB Perspectives - Laura's presentation:

  • It would be beneficial for users to be able to access datasets that are complementary without having to go from one DAAC to another to find the datasets.
  • Q: What would 'uniform metadata' look like?
  • Challenges:
    • Despite disparities in data access and resources, a challenge we all need to overcome is how to educate communities on what is available/possible based on the resources they do have
    • Re: merged products -they are easier to use for beginners; Copernicus (ESA) already makes a great merged chlorophyll product, do we need to make our own or should we establish connections to other agency resources to give users a broader range of resources from existing efforts across agencies
    • User story: Users typically want to work with an entire map of a specific product (like chlorophyll) or a time series of a specific region
    • (continuation of cross-agency linking) - linking to other agencies' tools and building relationships would also allow NASA to remain faithful to its' mission without feeling responsible for attempting to tackle and achieve all possible avenues of supporting users from just our agencies resources.


ESDIS Perspectives - Sandra's presentation:

  • Earthdata Search Client is the one-stop-shop for finding data from any NASA archive center and can support new user needs
  • ARSET is available for helping train users


Session 2 - OB.DAAC Updates


PACE Mission - Jeremy's presentation:

  • PACE mission reprocessing will occur in the coming weeks. Data version will increment, detailed release notes will be provided
  • NOTE: PACE data (and other mission data) goes through a scientific review board acceptance process that moves it between Test, Provisional, and Standard. PACE data are available to the public, but user should be aware that PACE data are currently at the PROVISIONAL stage.
  • UWG requests some form of notification (email, pop-up on dataset access) to any dataset they have historically downloaded so they can stay abreast of version updates/current data version. This enables consideration of methodology used to generate the data, and how it impacts products/algorithms that use the data (as well as associated uncertainty). Version updates may otherwise be overlooked as researchers may use data they already have to complete research. (NSIDC has a notification mechanism in place)
  • Communication as data changes is just as important as communication for new mission milestones like launch, first light, initial data release events. How can we keep the same momentum around communication as mission ages? * request repeated for SeaBASS data


Data Tools and Discovery - Alicia's presentation:

  • DAAC is focusing on moving toward enabling our historical processes through Earthdata tools.
  • Capabilities traditionally available in Level 1 & 2 browser are being implemented in Earthdata search. This will enable ordering through the EDSC interface, on the backend, OBPG still receives and processes the order and emails user to pick up staged files upon order completion.
  • Updates to OB.DAAC data tool interfaces were covered.
  • UWG mentioned the EO Browser (https://apps.sentinel-hub.com/eo-browser/) was a very good resource that uses a map to improve user ability to find data associated with products and/or regions (in comparison to EDSC interface). Useful features mentioned in EO Browser were scaling index, image based, ability to change bands.
  • UWG mentioned that the list of collections returned by EDSC tends to remind users of a 'google search' results page so when they are using the search bar or filters and the top few hits do not return expected datasets, they become confused.
  • UWG mentioned that it would be helpful to share user statistics with UWG and takeaways.


SeaBASS - Chris' presentation:

  • UWG inquired about the in-situ data QA/QC process and whether SeaBASS retroactively applies new QC standards to existing datasets and publishes it as a new version. New QC standards are not applied retroactively.


CALIOP Dataset - Kelsey's presentation:

http://orca.science.oregonstate.edu/lidar.data.php

  • UWG voted that OB.DAAC was the ideal place for CALIOP backscattering datasets to reside because 1) users would like to access these files in tandem with other ocean color datasets 2) having the data at OB.DAAC should accelerate discovery and use by the community
  • optimal formats for data need to be determined. Currently datasets are HDF5 tar files.


Algorithm Publication Tool (APT) - Brad's presentation:

  • APT can link into Earth and Space Journal if user desires submission
  • (Q:) How is CMR relating APT documents to other datasets?
  • UWG member shared that notification from APT interface wasn’t sufficient regarding changes to the document by other authors and progression of the document through review and publication stages.
    • Brad shared that feedback on changes to documents can be viewed by collaborators when they are logged into tool
  • flagging goes into ATBD


UWG Discussion - Day 1

  • Significant concern with the move to the cloud, UWG fears NASA is facing a HUGE RISK regarding equity since 1) universities can't afford the cost under current internal cost structures 2) grants are an avenue, but what happens when the grant ends? 3) who buys credits in the classroom/university setting?
  • (Concern) States, University systems, etc. have budget variations that need to be considered which was expressed by UWG as a serious, deep concern about equity.
  • (Suggestion) performing data transformation may help alleviate data size issues and barriers to machine capacity issues related to downloading data
  • partnerships with NIH and NSF are key to voicing concerns in unison
  • users like the ability to use Google Earth Engine (GEE) which allows processing online for free, but they are now worried that Google will start imposing a charge since other models have only provided free services temporarily. This greatly affects under-resourced users who rely on free services
  • UWG pointed out that having use cases defined for a variety of scenarios would help users understand the cost of cloud computing. They would have a better idea of a cloud computing provider/facility/resource.
  • As an example of what can go wrong with using cloud resources: UC-Berkley had data in Google and when the free tier ended abruptly, users were made to pay for data after 1GB.
  • With these concerns UWG wondered if they should advise students to use cloud or on-premise data
  • No labels
Write a comment...