IMPACT team members Muthukumaran R, Iksha Gurung and Aaron Kaulfus along with Anirudh Koul of SpaceML are convening a session at the American Geophysical Union #AGU21 fall conference that is focused on leveraging deep generative models and self-supervised learning techniques. The session will feature presentations that demonstrate how these methods are being leveraged to advance Earth sciences.

Presentation abstracts can be submitted here. More details are shown on the session poster below.

More information about IMPACT can be found at NASA Earthdata and the IMPACT project website.

The ongoing evolution in data, software, and computers is transforming the core tools of science and changing both what questions scientists ask and how they find the answers. IMPACT team scientist Dr. Chelle Gentemann is the first author of the research article “Science Storms the Cloud”, published in AGU Advances, which examines these trends and their implications for open science. We sat down with Dr. Gentemann to get her insights into the central themes of the article.

For decades, the method by which scientists worked with data was to download and analyze the data on their local computers. This imposed…

Satellite data is likely the first thing that comes to mind when most people think about NASA Earth observation data. Far fewer people know that NASA also relies on a host of other platforms for measuring Earth’s characteristics and for performing important science research. These include airborne (planes, drones, rockets, balloons), ocean (ships, boats, buoys and recently saildrones), or ground-based platforms (vehicles, trailers, observation towers, temporary field stations). …

In collaboration with the IEEE GRSS Earth Science Informatics Technical Committee, IMPACT held a machine learning summer school session for twenty-six global and diverse students. The session on “Scaling Machine Learning for Remote Sensing on Cloud Computing Environment” took place during the 4-day IEEE-GRSS event hosted by the working group on High-performance and Disruptive Computing in Remote Sensing (HDCRS).

The learning doesn’t stop in summer.

The goals of the session were multifaceted and included a focus on:

  • providing technical guidance on performing an end-to-end machine learning use case for remote sensing;
  • utilizing cloud computing for machine learning on remote sensing;
  • promoting open science via collaboration;
  • developing…

Given that a large haystack can contain over 7,000 cubic feet of hay, the aptness of the needle-in-the-haystack cliche becomes clear. Now imagine finding that proverbial needle when searching across the 139.7 million square miles of ocean that covers our planet. This was the task facing the IMPACT team working on an automated marine debris detection solution using machine learning.

An estimated eight million tons of plastic waste enters the ocean every year due to a lack of proper waste management. Runoff from land and river outflows lead to the accumulation of debris in coastal areas while ocean currents are…

IMPACT has released the open source, web-based tool ImageLabeler that allows users to create tagged images for use in training image-based machine learning (ML) models for Earth science phenomena. Training data for Earth science is as scarce as it is essential. One of the most challenging parts of scientific research is the collection of event cases for both detailed scientific investigation and climatological trend analysis. The ImageLabeler is designed to serve two purposes: 1) to provide a catalog of candidate events for scientific investigation, and 2) to gather, in a central location, the training data required to train machine learning…

Calling all all coders, scientists, entrepreneurs, designers, storytellers, makers, builders, artists, technologists, and space enthusiasts! NASA, ESA (European Space Agency), and JAXA (Japan Aerospace Exploration Agency) are hosting the Earth Observation Dashboard Hackathon from June 23–29.

The hackathon builds on the success of the the Space Apps COVID-19 Challenge. Over the course of a week, virtual teams will focus on utilizing data from the Earth Observing Dashboard to address challenges related to the COVID-19 pandemic. The teams will interact with experts from NASA, ESA, and JAXA as they develop their project submissions. The winning teams will have the opportunity to incorporate their solutions into the Earth Observing Dashboard.

Registration for the hackathon begins on May 20, 2021. Details are available here. Additional background information is available on NASA Earthdata.

The science of informatics is a big part of what IMPACT does within the field of the Earth sciences. But what is informatics? Is it just another word for information technology? Is it about information processing? Is it simply a focus on information systems or information science? A review of the literature on informatics easily returns any of the above concepts. But what does informatics mean practically as IMPACT implements it to benefit Earth science researchers?

The best way to demonstrate IMPACT’s approach to informatics is to allow the members of the informatics team to speak for themselves. We asked…

Machine learning can use human-labeled datasets as training datasets to achieve impressive results. However, hard problems exist in domains with sparse amounts of labeled data, such as in Earth science. Self-supervised learning (SSL) is a method designed to address this challenge. Using clever tricks that range from representation clustering to random transform comparisons, self-supervised learning for computer vision is a growing area of machine learning whose goal is simple: learn meaningful vector representations of images without having human labels associated with each image such that similar images have similar vector representations.

In particular, remote sensing is characterized by a huge…

Flood events result in devastating consequences for people, ecosystems, and economies. Floods occur throughout the world, and a single major flood event can result in multiple billions of dollars in damages. Flood extent is difficult to acquire on the ground as it is hazardous to operate in a flood zone, and physical access by roads is limited. Being able to detect flood extent provides automated ways to assist disaster response during flood events. Accurately detecting floods and flood extent via remote means greatly aids in the process of mitigating and responding to these destructive events.

Flooding caused by Hurricane Katrina

IMPACT, in collaboration with the…

IMPACT Unofficial

This is the unofficial blog of the Interagency Implementation and Advanced Concepts Team.

Get the Medium app

A button that says 'Download on the App Store', and if clicked it will lead you to the iOS App store
A button that says 'Get it on, Google Play', and if clicked it will lead you to the Google Play store