Carbon Analytics at Descartes Labs

An overview of Carbon Analytics terminology, products, and common interfaces.

Introduction

The Carbon Analytics API allows carbon project managers, accounting firms, and asset owners to accurately track the relevant carbon pools at their sites of interest around the globe, including critical changes through time.

This service is designed to provide maximum insight into carbon sequestered at a location, as well as the components that make up the total carbon pool. There are three carbon pools the Carbon Analytics service reports on:

Aboveground Biomass Carbon (AGB)

We use biomass data taken from the NASA GEDI L4A dataset to train our proprietary deep learning models. GEDI is a spaceborne LiDAR mission with a discontinuous collection of paths across the globe. We incorporate optical datasets as predictors to generate AGB estimates between collection tracks and for other points in time. 

Belowground Biomass Carbon (BGB)

Typically practitioners utilize country level (Tier 1) estimates for BGB which come from fixed AGB/BGB ratios. The Carbon Analytics service has implemented a peer reviewed technique that differentiates ratio values by geography and biophysical characteristics of the location. Additionally, the AGB component comes from the high resolution product we produce in this same Carbon Analytics service, resulting in a self-consistent data set update annually. The end result is a metric that aligns with Tier 3 methodologies.

Soil Organic Carbon (SOC)

Although not the primary component typically accounted for in forest carbon projects, soil organic carbon is still a quantity that may need to be tracked in support of sequestration initiatives. In support of this requirement, the Carbon Analytics service provides a blended soil data set from several global sources to provide a temporally static estimate of this quantity on an asset basis. Future improvements on this data set will leverage localized data, if available, to add better spatiotemporal characterization.

Product Components

The Carbon Analytics service consists of three individual components which provide carbon pool information specific to their use cases.

Carbon Time Series 

A core component of carbon accounting and credit generation is the ability to look back in time. The Carbon Time Series component allows customers to submit an asset polygon and retrieve annual changes to the carbon stock information going back twenty years. The carbon changes within the polygon are aggregated to provide a total annual number metric for the asset as a whole.

Carbon Raster Service 

Users may need to identity where carbon accumulation or loss is occurring within the asset as well. The Carbon Raster service allows users to retrieve asset specific raster data for each year in a twenty year period enabling analysis of spatial variability within the polygon.

Carbon Image Tile Service

The Carbon Image Tile Service allows users to request a set of carbon image tiles which can be easily integrated into a web mapping client of choice. The tile service is limited to most recent year and is available for all regions, not on the asset-scale. 

User Interfaces

There are two interfaces through which the user can query and retrieve Carbon Analytics data. 

Explorer

Subscribers to Carbon Analytics may load and create their own asset polygons through Descartes Labs Explorer. Once loaded, click the "Analyze" tab and use the "Drawing Tools" to either upload or draw a polygon up to 2000km².

• Select your "Area of Interest" polygon alongside a carbon pool "Dataset" and year of interest. This will load the Carbon Raster over your asset:

• Clicking "Show Carbon Timeline" will generate a Time Series of your asset polygon, where you can export as a csv directly through the component:

• Selecting a "Map Layer" will add the Tile Service as a background layer:

REST API

Please visit the API Docs for more details on accessing the Carbon Analytics API directly.