Skip to article frontmatterSkip to article content
Site not loading correctly?

This may be due to an incorrect BASE_URL configuration. See the MyST Documentation for reference.

Direct AWS S3 Access

S3 Access Overview

All NISAR data is hosted in NASA’s Earthdata Cloud (EDC), which leverages Amazon Web Services (AWS) infrastructure. As such, AWS Simple Storage Service (S3) access methods can be used to interact with NISAR datasets.

Data users can request temporary AWS credentials enabling direct access to the data in S3. This supports the use of a wide range of S3-aware tooling for interacting with cloud-hosted data as well as for low-latency and high-throughput data access patterns.

NISAR S3 Buckets

There are three S3 buckets associated with the NISAR mission:

Using the AWS CLI to Access NISAR Data

Users can leverage direct S3 access from other AWS services, such as EC2 and Lambda, as long as the resources are in the us-west-2 (Oregon) region.

1. Obtain your temporary AWS Credentials

You can either visit the NISAR S3 credentials endpoint directly, or use earthacess to programmatically retrieve your temporary credentials.

1.a. Visit https://nisar.asf.earthdatacloud.nasa.gov/s3credentials

If prompted, sign in with your Earthdata Login (EDL) credentials to retrieve a set of temporary AWS credentials, which will allow you to list and download contents of the S3 bucket. If you are already signed in, the credentials will display immediately.

The text displayed at the site provides four pieces of information:

For example:

{
 accessKeyId: "ASIAIOSFODNN7EXAMPLE",
 secretAccessKey: "wJalrXUtnFEMI/K7MDENG/bPxRfiCYEXAMPLEKEY",
 sessionToken: "LONGSTRINGOFCHARACTERS...6XUxCYEwbjGVKkzSNQh/",
 expiration: "2026-01-27 00:50:09+00:00"
}

More details about requesting temporary S3 credentials are available here.

1.b. Use the earthaccess Python package

import earthaccess

auth = earthaccess.login()

endpoint = 'https://nisar.asf.earthdatacloud.nasa.gov/s3credentials'
s3_credentials = auth.get_s3_credentials(endpoint=endpoint)

which will print, for example:

{
 "accessKeyId": "ASIAIOSFODNN7EXAMPLE",
 "secretAccessKey": "wJalrXUtnFEMI/K7MDENG/bPxRfiCYEXAMPLEKEY",
 "sessionToken": "LONGSTRINGOFCHARACTERS...6XUxCYEwbjGVKkzSNQh/",
 "expiration": "2026-01-27 00:50:09+00:00",
}

For more information about using the earthaccess package with NISAR, see the Earthaccess NISAR guide.

2. Configure your environment to use the temporary AWS credentials

You will need to set the environment variables in your terminal.

Linux or Mac users can export their credentials using the following commands:

$ export AWS_ACCESS_KEY_ID=ASIAIOSFODNN7EXAMPLE
$ export AWS_SECRET_ACCESS_KEY=wJalrXUtnFEMI/K7MDENG/bPxRfiCYEXAMPLEKEY
$ export AWS_SESSION_TOKEN=AQoDYXdzEJr...<remainder of session token>

Windows users can export their credentials using the following commands:

C:\> SET AWS_ACCESS_KEY_ID=ASIAIOSFODNN7EXAMPLE
C:\> SET AWS_SECRET_ACCESS_KEY=wJalrXUtnFEMI/K7MDENG/bPxRfiCYEXAMPLEKEY
C:\> SET AWS_SESSION_TOKEN=AQoDYXdzEJr...<remainder of token> 

For a full list of exportable variables, see AWS’s Temporary Credentials User Guide.

3. Find NISAR products of interest

NISAR production data is all hosted in the sds-n-cumulus-prod-nisar-products S3 bucket. Each product type has a prefix, and while you cannot list the full bucket contents, you can list the contents of each prefix. Refer to Prefix Structure for more information about the organization of the NISAR bucket and a table of prefixes for the NISAR data products and supporting products.

For example, if you want to list all available GCOV products, you can use this aws s3 ls command:

aws s3 ls s3://sds-n-cumulus-prod-nisar-products/NISAR_L2_GCOV_BETA_V1/

Within each product type prefix, there is a nested prefix for each individual product, with no additional hierarchy for grouping the individual products. Refer to Finding S3 Paths for tools and methods available to help you find specific products of interest.

Note that NISAR browse imagery and NISAR Urgent Response products are hosted in different buckets than the production NISAR datasets.

4. Open or download data files

Once you have an S3 path for an item of interest, you can download the file to your compute environment using the aws s3 cp AWS CLI command. You will need to include all relevant prefixes in your path. For example:

aws s3 cp s3://sds-n-cumulus-prod-nisar-products/PRODUCT_TYPE_PREFIX/PRODUCT_NAME_PREFIX/PRODUCT_NAME.ext path/to/local/dir

This example shows how to download one of the SME2 files from the S3 archive:

aws s3 cp s3://sds-n-cumulus-prod-nisar-products/NISAR_L3_SME2_BETA_V1/NISAR_L3_PR_SME2_008_007_D_073_4005_DHDH_A_20251216T164223_20251216T164300_X05007_N_F_J_001/NISAR_L3_PR_SME2_008_007_D_073_4005_DHDH_A_20251216T164223_20251216T164300_X05007_N_F_J_001.h5 .

Prefix Structure

Products are stored with the same prefix structure in all three NISAR S3 buckets. Each product type has a prefix, then each available product of that product type has an individual product prefix nested inside the product type prefix.

To access a product, the S3 path must contain the NISAR bucket name, the product type prefix, and the individual product prefix, followed by the full filename (including the extension), as illustrated here:
s3://sds-n-cumulus-prod-nisar-BUCKET/PRODUCT_TYPE_PREFIX/PRODUCT_NAME_PREFIX/PRODUCT_NAME.ext

Table 1 lists the prefix for each of the NISAR data product types, and Table 2 lists the prefix for each of the NISAR supporting data types.

Prefix Tables

Table 1:NISAR Data Product S3 Prefix List

ProductS3 Prefix
SME2NISAR_L3_SME2_BETA_V1/
GCOVNISAR_L2_GCOV_BETA_V1/
GUNWNISAR_L2_GUNW_BETA_V1/
GOFFNISAR_L2_GOFF_BETA_V1/
GSLCNISAR_L2_GSLC_BETA_V1/
RUNWNISAR_L1_RUNW_BETA_V1/
RIFGNISAR_L1_RIFG_BETA_V1/
ROFFNISAR_L1_ROFF_BETA_V1/
RSLCNISAR_L1_RSLC_BETA_V1/

Table 2:NISAR Supporting Data S3 Prefix List

ProductS3 Prefix
DEMDEM/

Production NISAR Data

NISAR data products are hosted in the sds-n-cumulus-prod-nisar-products S3 bucket. This bucket cannot currently be searched from the root, but you can list the contents from each product type prefix onward through the prefix hierarchy.

Here is an example of the S3 path for one of the available GCOV products:

s3://sds-n-cumulus-prod-nisar-products/NISAR_L2_GCOV_BETA_V1/NISAR_L2_PR_GCOV_005_172_A_008_2005_DHDH_A_20251122T024618_20251122T024652_X05007_N_F_J_001/NISAR_L2_PR_GCOV_005_172_A_008_2005_DHDH_A_20251122T024618_20251122T024652_X05007_N_F_J_001.h5

NISAR Browse Imagery

NISAR browse imagery is hosted in a different bucket than the actual data products. The prefix structure for sds-n-cumulus-prod-nisar-browse is the same as for the production NISAR data, but this bucket can be searched from the root directory if desired.

Within each product name prefix, there are two different browse and thumbnail files available in PNG format. The images displayed on the maps in Vertex and Earthdata Search are the main PNG files. Files tagged with _thumbnail.png are very small images used for applications like the Vertex search results, which require low resolution (100 x 100 pixels).

Here is an example of the S3 path for one of the available GCOV browse images:

s3://sds-n-cumulus-prod-nisar-browse/NISAR_L2_GCOV_BETA_V1/NISAR_L2_PR_GCOV_005_172_A_008_2005_DHDH_A_20251122T024618_20251122T024652_X05007_N_F_J_001/NISAR_L2_PR_GCOV_005_172_A_008_2005_DHDH_A_20251122T024618_20251122T024652_X05007_N_F_J_001.png

NISAR Urgent Response

The NISAR mission will produce Urgent Response (UR) products when natural disasters or other major events occur. These products may use less precise orbits in order to make data available as quickly as possible. UR products will be hosted in sds-n-cumulus-prod-nisar-ur-products, and will be retained for 30 days, allowing time for a standard product to become available before the UR product is deleted.

Finding S3 Paths

As more data becomes available, you may want to use wildcard searches with the aws s3 ls command to find specific products of interest. The product filenames contain a lot of information that can be used to restrict your search, such as track and/or frame numbers, acquisition dates, processing version, beam mode, polarization, and orbit direction.

Understanding the NISAR Naming Conventions can be helpful when constructing wildcard searches to narrow your results to specific products.

S3 Paths from Vertex

When you search for NISAR data products in Vertex, you can simply copy the S3 URL from the list of files for a search result, as shown in Figure 1.

Screenshot showing how to copy an S3 URL from Vertex search results

Figure 1:S3 URL links are available for files listed in Vertex search results for NISAR products.

To copy S3 URLs from Vertex:

  1. Search for the desired NISAR datasets in Vertex

  2. Select an item in the left panel of the search results

    • The files available for that item are listed in the right panel

    • You may need to adjust your browser window size to see the third panel

  3. In the right panel, click on the link icon next to the desired file

  4. Select S3 URL from the drop-down menu to copy the path to your clipboard

When you search for NISAR data products in Earthdata Search, S3 links are available to copy from the search results, as shown in Figure 2.

Screenshot showing how to copy an S3 URL from Earthdata Search results

Figure 2:S3 URL links are available for files listed in Earthdata Search results for NISAR products.

To copy S3 URLs from Earthdata Search:

  1. Search for the desired NISAR datasets in Earthdata Search

  2. Click on the Download button for the desired product

  3. Select the AWS S3 Access tab

  4. Click the filename to copy the S3 URL to your clipboard

S3 Access Limitations

Please note that direct S3 access for NISAR data is subject to these limitations: