Data Processing

NEON measures a diverse suite of biological, physical, chemical and ecological characteristics at field sites across the continent. NEON data are sent to headquarters after site construction is complete and data collection begins. The Observatory processes these measurements to derive standard, quality-assured data products that support greater understanding of complex ecological processes at local, regional and continental scales. Available NEON data, supporting metadata, science designs, data collection documentation and data processing documentation are accessible through the NEON Data Portal .

As shown in the example above, NEON processes raw field observations and sensor measurements to produce data products of scientific interest to the community. NEON divides data products and processing methods into three levels:

Raw data

Raw data are unprocessed measurements and observations from a single instrument, observation or field sampling techniques in native collection units, such as voltage.

Calibrated data

Calibrated or quality-assured data are generally from a single instrument, observer or field sampling area. These data are transformed into physical units, such as temperature, wind speed or radiance, and are generally at native measurement resolution. Calibrated data are used to create high-level, derived data products. Many sensors contain firmware that process the raw data into physical units, so calibration occurs within the sensor and NEON collects calibrated data from that sensor.

Derived data

Derived data products are 1) summarized individual measurements or 2) composed of multiple integrated measurements. NEON processes data to transform raw or calibrated data into biogeochemical units, or measures of environmental gradients that are not directly measured, such as Leaf Area Index (LAI). Derived products are often available at multiple space and time scales to capture seasonal and annual variability observed within NEON sites. The spatial coverage of derived data range from in situ and tower measurements to continental-scale data. Types of derived data products include:

  • Individual measurements that are further processed and summarized across a particular spatial resolution or time scale. For example:
    • Temperature or stream discharge measurements summarized for a daily, weekly or monthly average; or
    • Spectral reflectance data derived into a vegetation greenness and health (NDVI) product that covers an entire NEON site.
  • Several measurements incorporated to characterize ecological processes and change over time. These products sometimes incorporate airborne remote sensing data or data from multiple instruments. For example:
    • Vegetation density or Leaf Area Index (LAI) values that incorporate ground and airborne remote sensing measurements, processed to cover entire NEON sites; and
    • Soil respiration that is calculated using measurements of soil moisture, soil temperature and carbon dioxide gas concentration, as well as soil properties like permeability.
  • Continental-scale, derived data products often utilize spatial models that incorporate understanding between ecological drivers and responses gained at NEON sites and then use external datasets that characterize such drivers to predict conditions in areas where NEON does not collect data. These models produce datasets that characterize ecological processes and conditions continuously across the entire continent. These data products, coupled with other NEON datasets, facilitate understanding of long-term ecological change and seasonal or annual variability in at regional scales where they have large impacts. For example:
    • Mathematical and statistical models can be used to predict water evaporation to the atmosphere, the amount of carbon stored in vegetation and soil moisture across the continent using information about the weather and land cover, even where NEON does not directly collect data.

Data products selection criteria

NEON data products are strategically selected based on the following criteria:

  • Identified needs of the research community
  • Usability of the data
  • Ability to effectively characterize the causes and effects of ecosystem-level change
  • Relevance in predictive modeling and forecasting of future ecosystem conditions and states

Browse NEON data

Dialog content.