Skip to content

← Back

Biomass (Above and Belowground) (v3.0c, 3.1g)

Using above and belowground biomass data, SOCSPOT 3.0c and 3.1g estimate long-term soil organic carbon sequestration potential, enhancing global models and offering insights into plant resilience to drought.

To access high-resolution data for your area(s) of interest, contact our team.

High-Level Description

SOCSPOT 3.0c and 3.1g incorporate incoming carbon from both above and belowground biomass in order to better understand long-term (20-30 year) relative soil organic carbon (SOC) sequestration potentials. The above and belowground biomass carbon inputs leverage understanding of interactions among the plant, decomposition, and rhizosphere in order to best estimate the amount of incoming carbon at the highest resolution possible, globally, across rangelands and croplands. This product relies on remotely-sensed net primary productivity, landcover identification, and phenological allometry in order to estimate belowground biomass.

Use Cases

With above and belowground biomass, customers can:

  • Improve global models reliant upon high-resolution and quality biomass inputs.
  • Better understand plant resilience to drought (greater rooting mass/depths can increase resilience).

Technical Specifications

  • Native Resolution: 10-30m
  • Spatial Extent: CONUS(3.0c), Global(3.1g)
  • Domain: Croplands and rangelands (annual and perennial grasslands, trees, shrub and scrub)
  • Units: tonnes C / ha / yr
  • Inputs: Aboveground biomass, annual vegetation type
  • Temporal Range: 2018-2022

Algorithm Theoretical Basis

SOCSPOT 3.0c and 3.1g fuse remotely-sensed aboveground biomass estimates, landcover identification, and empirical relationships from phenological allometry in order to estimate total incoming carbon for each pixel. For both crop and rangeland, the aboveground biomass is combined with allometric equations specific to each crop or plant functional group to estimate the belowground biomass. Within crop-specific land covers, the proportion of harvested versus residual material is also accounted for before estimating potential C-inputs from aboveground biomass.

Model Assumptions & Constraints

A number of assumptions and methodological choices are important to consider when using outputs of SOCSPOT3.1g.

  • For croplands, we assume that the carbon left over in residues (i.e., residual biomass following harvest) is incorporated into the soil rather than being transported off farm or burned.
  • Global crop-type identification only exists for one year (2021) and does not include distinctions beyond corn, wheat, and ‘other crop’. For ‘other crop’, generalized crop coefficients are used for the allometric equations and we assume that land parcels are consistently growing the same crop.
  • While vegetation and land cover can change throughout the year we use the vegetation cover that exists most commonly (i.e., the mode) within a pixel throughout a given year. Because the default resolution (500m) is coarser than the land cover underlying these pixels, the mode is not only temporal, but also spatial (within each 500m pixel).
  • For v3.0c, distinct methodologies exist for rangelands and croplands.
  • Rangelands use RAP for both fractionated vegetation cover and biomass.
  • Croplands use the CDL for identifying vegetation type and MODIS for biomass inputs.
  • Croplands also use the CDL to establish crop residuals remaining on each field. However, unique residuals are not included for every crop type within the CDL. In these cases, we assume a generic residual coefficient for crops with less information available.

Known Issues

  • Within-year crop rotations and/or interspersed multi-cropping systems are modeled as a single crop due to constraints on satellite identification of such land parcels and limited understanding of carbon cycling within such systems.
  • Within v3.1g, different landcover vegetation types are identified at different portions of the year, and inter-annually. As such, we assume the most common (i.e., statistical mode) land cover type within the year is the predominant vegetation type.
  • Currently, Dynamic World and Global Pasture Watch are used to identify potential rangelands based on specific land covers, which include the label of ‘potential’ because they may not be grazed. Until global coverage of grazed land area is available, this label will remain as ‘potential rangelands’.

What’s New?

With the introduction of SOCSPOT 3.1g, we add:

  • With the introduction of SOCSPOT 3.1g, we expand to include global coverage of biomass carbon inputs and switch to using MODIS across both rangelands and croplands.
  • With the introduction of SOCSPOT 3.1g, novel globally-gridded landcover dataset that reflects a fusion of multiple landcover data sources.
  • Version 3.0c is the first to introduce crop-specific residue estimates for aboveground biomass as well as allometric estimates of belowground biomass within Continental United States (CONUS).
  • Version 3.0c introduces mineralization adjustments for belowground biomass carbon inputs for both rangelands and croplands.
  • Version 3.0c uses CDL and RAP to identify LULC classes, while 3.0g uses a fused product including Dynamic World, Global Pasture Watch, and ESA World Cereals.

Data Partners, Providers, and References

  • NASA Land Processes Distributed Active Archive Center (LP DAAC), USGS/Earth Resources Observation and Science (EROS) Center: Running, S.W., Zhao, M., & Nemani, R. 2015. MOD17A3HGF MODIS/Terra Net Primary Production Yearly L4 Global 500m SIN Grid V006. NASA EOSDIS Land Processes DAAC. DOI: 10.5067/MODIS/MOD17A3HGF.006.
  • Rangeland Analysis Platform: Allred, B.W., Smith, W.K., Twidwell, D., et al. 2021. Rangeland Analysis Platform: A user-friendly online tool for monitoring rangelands. Rangeland Ecology & Management, 74, 36-42. DOI: 10.1016/j.rama.2020.11.009.
  • USDA National Agricultural Statistics Service: Cropland Data Layer. USDA NASS. 2022. Cropland Data Layer. Published crop-specific land cover data. Washington, DC: USDA National Agricultural Statistics Service.
  • Brown, M., Perger, C., & Esch, T. 2022. Dynamic World: Near real-time global 10 m land use land cover. Google Earth Engine. DOI: 10.1038/s41597-022-01407-z.
  • ESA and Partners: ESA WorldCereal Project. 2021. WorldCereal: A global-scale mapping of cropland and crop types.
  • Pasture Watch, Land and Carbon Lab, Satelligence. 2023. Global Pasture Watch: Pasture Monitoring Platform.

Accessing Data

Contact our team to request high-resolution data for your area(s) of interest.