Skip to content

Jack Hayes final project for Winter 2025. Evaluating ICESat-2 ATL06 algorithms over varying land cover and terrain

License

Notifications You must be signed in to change notification settings

UW-GDA/ATL06_LaserLottery

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

21 Commits
 
 
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

ATL06: Laser Lottery what da dog doin

Project Team

  • Jack Hayes (@Jack-Hayes)
  • Collaborators (this could be you!)

Project Overview

ATL06: Laser Lottery is a comparative analysis of ICESat-2's ATL06 land ice elevation algorithm and custom photon-counting approaches. Using a variety of elevation datasets, we aim to assess the impact of different processing strategies on elevation retrievals over diverse terrain and vegetation. Our goal is to determine why certain processing algorithms perform better in specific conditions and how they can contribute to NASA's Surface Topography and Vegetation (STV) Incubation program efforts.

Background

Launched on September 15, 2018, ICESat-2 is a NASA satellite that measures global elevation using the Advanced Topographic Laser Altimeter System (ATLAS). ATLAS fires 10,000 laser pulses per second, each containing ~300 trillion photons, but only a handful return to the sensor. These photons are then processed by algorithms such as ATL06, which estimates land ice surface height by filtering noise and computing along- and across-track slopes.

ICESat-2 Beams (Smith et al., 2019)
Figure: ICESat-2 beam configuration (Smith et al., 2019).

ICESat-2 supports multiple applications through various data products, with a few listed below:

  • ATL03: Raw photon cloud data
  • ATL06: Land ice elevation (our focus)
  • ATL08: Canopy height and surface classification
  • ATL13: Inland water surface heights

For more details, see the ICESat-2 Data Products.

Problem Statement & Objectives

Accurate and high-resolution topographic data are crucial for understanding Earth’s dynamic surface processes, including glacier retreat, vegetation structure, and elevation changes. The National Academy of Sciences’ 2017–2027 Decadal Survey for Earth Observation identified these datasets as priority observables for NASA Earth Science (National Academies, 2018). In response, NASA’s Surface Topography and Vegetation (STV) Incubation Program aims to develop next-generation multi-modal elevation datasets by integrating satellite and airborne data.

Unlike past efforts such as SRTM (Shuttle Radar Topography Mission), which provided a near-global topographic dataset over 11 days in 2001, STV seeks to fuse elevation measurements from lidar, optical stereo, and synthetic aperture radar (SAR) collected at different times. This integration introduces several challenges:

  • Variability in measurement techniques (e.g., photon-counting lidar vs. radar)
  • Differences in spatial resolution and sensor geometry
  • Real surface changes over time (e.g., vegetation growth, ice melt)

Given the limited scope of our final project, we're mainly concerned with the question what biases are introduced by different ATL06 processing parameters, how do these biases change over varying terrain and land cover characteristics, and why?

ATL06 is a standard processing algorithm for ICESat-2, but its performance and biases in non-glacial environments remain poorly understood. Comparing customized ATL06 photon processing methods, using the "standard" ATL06 processing as reference, is essential to:

  • Identify systematic biases in ICESat-2’s elevation products
  • Understand why ATL06-derived elevations disagree with other altimetry sources
  • Improve the accuracy of fused elevation datasets for STV

In the context of our project, we will hone in on understanding why ATL06-derived elevations disagree with "ground truth" altimetry derived from aerial LiDAR DEMs over differing vegetation. These DEMs are genereated from the USGS 3DEP program and are commonly used as "ground truth" reference for elevation/altimetry comparison studies. We hope to gather diverse samples of ATL06 measurements across diverse vegetation types and ecozones over the contiguous US.We will evaluate the performance over various custom ATL06 processing algorithms over these sites. These sites have been chosen to eliminate temporal decorrelation between ICESat-2 and the reference DEMs, where ICESat-2 points are within 14 days of the USGS DEM acquisition date. Acquisition years range from 2019 to 2022.

Datasets

We will analyze ICESat-2 elevation data alongside:

Tools & Software

We will leverage multiple tools to process and analyze the data:

  • SlideRule – Cloud-based ICESat-2 processing framework, supporting both standard and custom algorithms.
  • GeoPandas – Used for spatial data analysis for our ICESat-2 points and other relevant vector geometries.
  • Xarray – Used for handling multi-dimensional elevation datasets and other relevant rasters and nDarrays.
  • easysnowdata – Used for grabbing NLCD data thanks to my favorite 5th year grad student 👉👈🥰.
  • probably scipy stats and numpy for elevation value comparison and evaluation

Methodology

  1. Preprocess Differing Elevation Measurement Sources – Retrieve elevation profiles over our predetermined sites. (maybe subsample the sites)
  2. Apply Custom Photon-Counting Algorithms – Implement alternative processing methods in SlideRule via gridsearch.
  3. Compare with Reference Datasets – Validate against USGS 3DEP DEMs.
  4. Tie Findings into NASA’s STV Framework – Assess how these techniques fit into broader Earth observation objectives.

Expected Outcomes

  • Identification of vegetation where custom photon-counting algorithms outperform standard ATL06 processing.
  • Insights into vegetation-dependent biases in ICESat-2 elevation retrievals.
  • Contributions to NASA’s STV Incubation program for multi-modal elevation fusion.

Related Work

References

  • National Academies of Sciences, Engineering, and Medicine (2018). Thriving on Our Changing Planet: A Decadal Strategy for Earth Observation from Space. DOI: 10.17226/24938.
  • Donnellan, A. (2021). Observing Earth's Surface Topography and Vegetation Structure in the Next Decade. AGU U32A-01.
  • Farr, T. (2007). Shuttle Radar Topography Mission (SRTM) Data Processing and Applications.

Because counting photons is more complicated than it sounds 🚀❄️

About

Jack Hayes final project for Winter 2025. Evaluating ICESat-2 ATL06 algorithms over varying land cover and terrain

Resources

License

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published