Accessing ICESat-2/ATL11 Land Ice Height time-series

In this tutorial, we’ll step through a data pipeline on turning point cloud time-series data from the ICESat-2 laser altimeter into an analysis ready format. By the end of this lesson, you should be able to:

• Access ICESat-2 data in a Hierarchical Data Format (HDF5) from the cloud

• Construct a data pipeline that combines multiple ICESat-2 tracks and laser beams into a flat data structure

• Store the pre-processed data in a cloud-optimized, analysis ready data format

References:

• https://github.com/weiji14/deepicedrain/blob/v0.4.2/atl06_to_atl11.ipynb

• https://github.com/weiji14/deepicedrain/blob/v0.4.2/atl11_play.ipynb

About ICESat-2

The Ice, Cloud, and land Elevation Satellite-2 (ICESat-2) was launched in 2018. The Advanced Topographic Laser Altimeter System, or ATLAS, is the only instrument on board. ATLAS has a single green laser that is split into six beams, arranged in three pairs. The 10,000 laser pulses emitted each second reach the earth and reflect off the surface before returning to the satellite, where their travel time is recorded and ultimately used (in combination with information about the satellite’s location) used to determine the height of the surface they reflected off of.

Data Products

The photon travel times collected by ATLAS are ultimately processed into a series of ICESat-2 data products. The data products are produced with multiple levels of processing, from geolocated photons (ATL03) to gridded time series (e.g. ATL11). For this analysis, we use one of the highest level (3B) products: ATL11 Slope-Corrected Land Ice Height Time Series product [Smith et al., 2021].

Data Access

ICESat-2 data access is available from NSIDC through several mechanisms, including for local download and in the cloud. A compilation of resources for accessing and working with ICESat-2 data is available in this resource guide and through the NSIDC website. Here we will access data in the cloud by getting the appropriate s3urls using icepyx, a Python software library and community of ICESat-2 data users, developers, and data managers.

Getting started

These are the tools you’ll need.

import datatree
import earthaccess
import icepyx as ipx
import xarray as xr

Just to make sure we’re on the same page, let’s check that we’ve got compatible versions installed.

xr.show_versions()

Cloud access to ICESat-2 ATL11 files

In this book we use icepyx for gathering the necessary s3 urls to access ICESat-2 data in the cloud. icepyx can also be used (with nearly identical syntax) to download data to your local machine. For more details on accessing ICESat-2 data in the cloud, please check out the references below!

References:

• https://github.com/icesat2py/icepyx/blob/development/doc/source/example_notebooks/IS2_cloud_data_access.ipynb

• https://nsidc.github.io/earthaccess/tutorials/demo/

• https://nasa-openscapes.github.io/earthdata-cloud-cookbook/examples/NSIDC/ICESat2-CMR-AWS-S3.html#data-access-using-aws-s3

• https://nsidc.org/data/user-resources/help-center/nasa-earthdata-cloud-data-access-guide

• https://book.cryointhecloud.com/tutorials/IS2_ATL15_surface_height_anomalies/IS2_ATL15_surface_height_anomalies.html

Providing credentials

Accessing NASA data requires you to have an Earthdata Login. You can sign up for one free at https://data.nsidc.earthdatacloud.nasa.gov/ and learn more about NASA authentication and managing your credentials via Earthaccess and here.

By obtaining your s3 urls via icepyx, you are also able to authenticate for cloud data access (note: Earthaccess is used under the hood to do this).

# First we must let icepyx know where (and when) we would like data from.

short_name = "ATL11"  # The data product we would like to query
spatial_extent = [-180.0, -85.0, 180.0, -60.0]  # bounding box for Antarctica
date_range = ["2018-09-15", "2023-05-31"]  # entire satellite record

# Setup the Query object
region = ipx.Query(short_name, spatial_extent, date_range)

region.visualize_spatial_extent()

# Get the granule IDs and cloud access urls (note that due to some missing ICESat-2 product metadata, icepyx is still working to provide s3 urls for some products)
gran_ids = region.avail_granules(ids=True, cloud=True)
print(len(gran_ids[0]))
print(gran_ids[0][:10])

# Authenticate using your NASA Earth Data login credentials; enter your user id and password when prompted
region.earthdata_login(s3token=True)

# set up our s3 file system using our credentials
fs_s3 = earthaccess.get_s3fs_session(daac="NSIDC", provider=region._s3login_credentials)

Loading into xarray

Let’s read a single ICESat-2 ATL11 HDF5 file into an xarray data structure!

First we’ll take a quick look at an example of an ATL11 HDF5 file. We’ll read it using xarray.open_dataset.

# s3_url = gran_ids[0][3]
s3_url = "s3://nsidc-cumulus-prod-protected/ATLAS/ATL11/005/2019/09/30/ATL11_005411_0315_005_03.h5"

with fs_s3.open(path=s3_url) as h5file:
    ds = xr.open_dataset(h5file, engine="h5netcdf")
ds

Hmm, so there are a bunch of attributes, but no data variables. This is because the ICESat-2 laser altimeter data is stored in ‘groups’ per laser.

For ATL11, the 6 lasers have been combined into 3 pair tracks (pt1, pt2, pt3). To read the nested data structure, we can either loop over each of these groups, and/or use something like datatree.open_datatree.

References:

• https://medium.com/pangeo/easy-ipcc-part-1-multi-model-datatree-469b87cf9114

with fs_s3.open(path=s3_url) as h5file:
    pair_track_dict = {}
    for pair_track in ["pt1", "pt2", "pt3"]:
        pair_track_dict[pair_track] = xr.open_dataset(
            filename_or_obj=h5file, engine="h5netcdf", group=pair_track
        )
    dt = datatree.DataTree.from_dict(d=pair_track_dict)
dt